key: cord-0006336-m9ax1hx5
authors: nan
title: Neurocritical Care Society 11th Annual Meeting: October 1-4, 2013 Philadelphia Marriott Downtown Hotel Philadelphia, Pennsylvania
date: 2013-09-17
journal: Neurocrit Care
DOI: 10.1007/s12028-013-9895-1
sha: 3a1be09d6b697d7ce6735ebbd641dcc4f0884b10
doc_id: 6336
cord_uid: m9ax1hx5

nan

CNS neuroinflammation can cause brain injury from direct neurotoxicity and/or mass effect. We recently demonstrated that astrocytic Transforming Growth Factor Beta (TGFbeta) signaling is required to limit neuroinflammation after ischemic stroke. We utilized Toxoplasma gondii to investigate whether astrocytes also utilize TGFbeta signaling to limit T lymphocyte responses during CNS infection.

We orally infected "DN" mice, which have impaired astrocytic TGFbeta signaling, and wildtype (WT) controls. We assessed brains early and late after Toxoplasma reached the brain, at 2 weeks and 4 weeks. We measured CNS Toxoplasma burden with quantitative PCR, and neuronal injury by immunohistochemisty for MAP-2 and NeuN. CNS inflammatory responses were assessed by immunohistochemisty for activated astrocytes (GFAP) and microglia (CD68), and T lymphocytes (CD3), stereology for T cell numbers, flow cytometry to examine T lymphocyte subsets, and multiplex luminex assay and ELISA to measure cytokines and chemokines in the brain.

WT and DN mice exhibited no differences in weight loss or Toxoplasma load. However, DN mice exhibited 2-fold more neuronal destruction than WT mice (P<0.05). This was associated with 2-3-fold more microgliosis and astrogliosis (P<0.05 and <0.01, respectively). The number of focal T cell infiltrates was the same in DN and WT mouse brains however DN mice exhibited more diffuse T lymphocyte infiltration and 2.5-fold higher overall numbers (P<0.01). Notably the 26 cytokines and chemokines measured by multiplex luminex assay demonstrated typical Th1 responses in both mutant and WT mice, and the only significant difference between genotypes was in the T-lymphocyte chemoattractant RANTES (156.5±34.57 pg/mg tissue in DN vs. 71.02±13.77 in WT, P<0.05).

Astrocytic TGFbeta signaling is required to limit RANTES expression during CNS Toxoplasma infection. It suppresses T cell migration into the brain and limits neurotoxicity. Astrocytic responses to TGFbeta may be a universal pathway by which the CNS limits neuroinflammation.

Financial Support: Stanford Institute for Immunity, Transplantation and Infection and NIH-NINDS (R01 to MSB, K08 to AAK)

In 1981 the Uniform Determination of Death Act (UDDA) defined brain death (BD) as the "irreversible cessation of entire brain function" and established its legitimacy as a mechanism of death. Despite widespread acceptance of BD in the United States, a pivotal study recently demonstrated significant variability in BD policies at major medical centers and their non-congruence with American Academy of Neurology (AAN) guidelines. We sought to characterized the present state of BD determination in actual practice.

We reviewed the charts of all adult (>16 years) BD organ donors during 2011 from 68 heterogeneous hospitals in the Midwest United States. Data was collected across five categories: guideline performance, pre-clinical testing, clinical examination, apnea testing, and use of ancillary tests. Practice within categories and overall adherence to AAN guidelines was assessed.

226 BD organ donors were included. Practice exceeded recommendations in guideline performance but varied widely and deviated from AAN guidelines in all other categories. Pre-requisites were frequently not satisfied prior to the clinical examination. 102 (45.1%) had complete documentation of brainstem areflexia and absent motor response. 166 (73.5%) had completed apnea testing. Of the 60 without completed apnea testing, 56 (93.3%) had ancillary tests consistent with BD. Overall, only 101 (44.7%) strictly and 84 (37.2%) loosely adhered to contemporary AAN guidelines.

1. There is wide variability in the documentation of BD determination, likely reflecting similar variability in practice. 2. With respect to clinical examination alone, the brain death examination is most often incompletely documented. 3. This is a call to action for improved documentation, better uniformity of policies and comprehensive and strategically targeted educational initiatives to ensure consistently contemporary approaches to BD determination in every patient. Neurocritical care specialists whould be the key advocates for promoting these necessary changes.

The quantity of subarachnoid (SAH) and intraventricular hemorrhage (IVH) occurring in the setting of a ruptured cerebral aneurysm is strongly associated with complications and poor outcomes.

We performed a double-blind, placebo-controlled, randomized pilot trial to assess use of intraventricular TPA in patients with aneurysmal SAH. Patients were eligible if they had a modified Fisher score of 4; underwent endovascular coil embolization; and had an external ventricular drain (EVD) in situ. Patients were allocated 1:1 in blocks of 4 to receive either 2 mg TPA every 12 hours (maximum 10 mg) or an identical placebo. CT scans were performed approximately 12, 48 and 72 hours and 8 days after drug administration. Primary outcomes included: (1) Feasibility (enrolment and consent rates); (2) Safety, assessed by prospectively screening for complications; and (3) Rate of intracranial blood clearance, assessed using SAH sum and IVH scores. Secondary outcomes included vasospasm (assessed by day 8 CT angiography), delayed cerebral ischemia (DCI), need for ventriculoperitoneal shunting and 6-month neurological outcomes.

Seventy-seven patients were screened, 16 met inclusion criteria, and 12 were randomized. The consent rate was 88%. There were no cases of new intracranial hemorrhage complicating use of TPA. EVD-related infections occurred in one patient per group (17%). The median percent reduction in IVH volume (measured in ml) versus baseline was larger in TPA-treated patients at 12 hours (73% vs. 10%, p=0.06) and 48 hours (94% vs. 17%, p=0.04), but not 72 hours (90% vs. 49%, p=0.42). Results were similar for SAH sum score (49% vs. 9%, p=0.04; 86% vs. 27%, p=0.02; and 88% vs. 43%, p=0.07, respectively). Among patients treated with TPA, SAH clearance was enhanced by earlier administration (β=-0.002, p=0.02). There was no difference in the proportion of patients developing vasospasm, DCI, need for permanent shunting or poor 6-month outcomes.

Intraventricular TPA accelerates clearance of SAH and IVH without increasing the rate of hemorrhagic complications. A larger-scale clinical trial of intraventricular thrombolysis is feasible, will need to be conducted at multiple centers, and is required to determine whether this practice reduces complications and improves neurological recovery.

Financial Support: Support from the Department of Critical Care Medicine, University of Calgary and Hotchkiss Brain Institute

Cortical spreading depolarization (CSD) induces dynamic changes in regional cerebral blood flow (rCBF) resulting in either transient hyperperfusion (physiological neurovascular coupling) or hypoperfusion (pathophysiologic inversion= spreading ischemia). The purpose of this study was to evaluate the Bowman Hemedex® rCBF monitor, a parenchymal thermal diffusion probe, to detect dynamic changes in rCBF in response to CSD in severe TBI patients.

Twelve adult TBI patients requiring craniotomy were enrolled. During surgery, an electrode strip was placed on peri-lesion cortex and a thermal diffusion rCBF probe in adjacent parenchyma,<1 cm from the strip. Electrocorticography and rCBF were recorded bedside in the neurointensive care unit for up to 7 days.

rCBF was obtained for 56% of the duration of electrocorticography, with an average rCBF of 39.6 ± 8.0 ml/100 g/min. A total of 477 CSDs were recorded in 8 patients (range 0-132 CSDs per patient) and simultaneous rCBF was obtained during 147 CSDs. Stereotyped rCBF responses, including both hyperperfusions and hypoperfusions, were observed in a time-locked fashion relative to CSDs in 60 of 147 cases. In one patient, consistent 40% transient decreases in rCBF occurred in association with 9 CSDs. Dynamic evolution of neurovascular coupling was also observed. One patient exhibited hyperperfusions initially (28% increase, n=13) followed by hypoperfusions with subsequent CSDs (30.4% decrease, n=13). Another patient exhibited a progressive worsening of spreading ischemia with successive CSDs: initial events produced a 19.8% decrease in rCBF (n=4), which later augmented to 34.2% decreases (n=7) and then to 56.2% decreases (n=14).

The Combined Approach to Lysis Utilizing Eptifibatide and rt-PA in Acute Ischemic Stroke-Enhanced Regimen (CLEAR-ER) trial showed that IV rt-PA plus eptifibatide was safe, and provided evidence that a phase III trial to determine efficacy of the regimen is warranted. We compared the rate of sICH and the proportion of good outcomes for patients in the CLEAR-ER trial to a matched cohort of patients from the original NINDS rt-PA trial.

CLEAR-ER was a multi-center, double-blind, randomized safety study. Ischemic stroke patients with NIHSS ≥ 6 and age ≤ 85 were randomized to 0.6mg/kg rt-PA plus eptifibatide (135mcg/kg bolus and a two-hour infusion at 0.75mcg/kg/min) versus standard rt-PA (0.9mg/kg). For this analysis, we matched 1:1 CLEAR-ER patients who received eptifibatide plus rt-PA with similar patients in the rt-PA arm of the NINDS trial and compared safety (sICH rate) and good outcomes (mRS ≤1 or return to baseline mRS). Patients were matched by age, gender, race, baseline mRS, baseline NIH stroke score, and stroke onset to rt-PA (minutes).

Fifty four matches were made. One (1.8%) sICH occurred in each group (odds ratio 1.00, 95%CI 0.01-78.50). At 90 days, 51.8% of the CLEAR-ER group had good outcomes versus 46.3% in the NINDS rt-PA group (odds ratio 1.30, ). For subjects with baseline NIHSSS>12 (CLEAR-ER median NIHSSS), 38.5% of the CLEAR-ER group had good outcomes versus 23.1% in the NINDS group (odds ratio 2.33, 95%CI 0.60-9.02).

We confirmed that eptifibatide plus rt-PA showed similar safety compared with rt-PA alone. Further, the direction of effect on outcomes favored the eptifibatide group although this study was not powered to detect differences in efficacy and the results were not statistically significant. A phase III trial is needed to explore the benefits of eptifibatide plus rt-PA for patients with acute ischemic stroke.

Financial Support: William Knight -Speaker's Bureau -Genentech, Inc Genentech, Inc provided rt-PA as study drug for the CLEAR-ER trial. Merck provided Eptifibitide as study drug for the CLEAR-ER trial The CLEAR-ER trial was funded by an NIH grant

Epsilon (e) variants in the Apolipoprotein E (APOE) gene are associated with lobar and non-lobar spontaneous intracerebral hemorrhage (s-ICH). We aimed to test the hypothesis that APOE variants are also associated with warfarin-related ICH (w-ICH) and examine the interaction between APOE and warfarin exposure in causing ICH.

In this prospective multicenter study, ICH was classified as lobar and non-lobar based on admission head computed tomography. W-ICH cases were matched with warfarin-exposed controls (w-controls), randomly selected from ambulatory clinics. APOE-e variants were directly genotyped. A case-control design (w-ICH cases and w-controls) was utilized to test for association between APOE-e variants and w-ICH, implementing logistic regression for hypothesis testing. A case-only design (s-ICH and w-ICH cases) was utilized to test for interaction between APOE variants and warfarin, implementing logistic regression using warfarin and APOE status to test for gene-environment interaction.

Although blocking NMDA-mediated excitotoxicity remains a promising strategy for neuroprotection in acute brain injury, the non-selective nature of most NMDAR antagonists has limited clinical translation due to toxic side effects. We now test the efficacy of NP-10679, a novel pH sensitive NMDAR antagonist designed to work selectively in areas of tissue acidosis associated with ischemia.

Following induction of experimental SAH by perforation of the right middle cerebral artery (MCA), 44 male C57-BL/6 mice were randomized into three treatment groups: vehicle; 2 mg/kg NP 10679; 5 mg/kg NP10679. Drug was administered intravenously by tail vein injection immediately after injury and every 12 hours for 3 days. Daily functional testing was performed by a blinded investigator, and included Rotorod latency as a measure of vestibulomotor deficit and Neuroseverity score as a measure of sensorimotor deficit. Animals were sacrificed at 72 hours, and perfused with gelatin/India Ink to assess degree of vasospasm in the right MCA.

All three groups received a similar hemorrhage as assessed by histological measurements after sacrifice. However, we observed a significant improvement in Rotorod and Neuroseverity scores in treated vs. untreated mice (p<0.01 in treated groups vs. vehicle group). The effect was dose dependent and persisted for 3 days of testing. Histological examination after ink-gelatin perfusion showed similar luminal diameters of basilar and contralateral MCA across groups; however, there was significantly less vasospasm in the right MCA adjacent to clot in both treated groups (p = 0.01, ANOVA) as compared to vehicle group.

Our results suggest that pH sensitive NMDA receptor antagonist improves functional recovery and reduces histological evidence of vasospasm in a validated murine model of subarachnoid hemorrhage. Selective NMDA receptor antagonists designed to work selectively in areas of tissue injury may be a promising strategy to prevent delayed cerebral ischemia following SAH.

Dantrolene is neuroprotective in animal models and may attenuate cerebral vasospasm (cVSP) after aneurysmal subarachnoid hemorrhage (aSAH) in humans. We evaluated safety/tolerability and feasibility of intravenous dantrolene (IV-D) after aSAH.

In this single-center, randomized, double-blind, placebo-controlled trial, we randomized 31 patients with aSAH to IV-D 1.25 mg IV every 6 hours x 7 days (n=16) or placebo (n=15). Primary endpoint was incidence of hyponatremia (sNa ≤ 134 mmol/L) and liver toxicity (% patients with ALT, AST and AlkPhos >5x upper limit of normal). Secondary safety endpoints included tolerability, systemic hypotension and intracranial hypertension. Efficacy assessments included clinical, transcranial Doppler (TCD) or angiographic cVSP occurrence, delayed cerebral ischemia (DCI) and 3-month modified-Rankin-Scale, Glasgow Outcome Scale and Barthel Index. Statistical analysis was performed using non-parametric tests, generalized estimating equations and mixed models.

Between IV-D vs. placebo, no differences were observed in the primary outcome (hyponatremia: 44% vs. 67% [p=0.29] ; liver toxicity 6% vs. 0% [p=1.0]), or in AEs or SAEs (16 vs. 5 AEs, of which 5 vs. 2 were severe; RR 2.2; 95% CI 0.7-6.7; p=0.16). Three IV-D vs. two placebo patients reached stop criteria: one IV-D patient developed liver toxicity; two patients in each group developed brain edema requiring osmotherapy. No differences in angiographic, TCD, clinical cVSP, DCI, or 3-month functional outcomes were seen. Quantitative angiogram analysis revealed a trend towards larger vessel diameters in the IV-D group after the 7-day infusion-period (p=0.05).

IV-Dantrolene after aSAH is feasible and safe. This small trial was underpowered to show efficacy or outcome differences.

Financial Support: This study was funded by the American Heart Association Founders Affiliate (09SDG2030022), the Worcester Research Foundation, and the University of Massachusetts Medical School (Faculty Scholar Award). This study was also supported by resources enabled by the Clinical Translational Science Award (5UL1TR000161) to the University of Massachusetts Medical School.

Background: Diffusion weighted imaging (DWI) abnormalities in the setting of intracerebral hemorrhage (ICH) has been suggested in some studies, and could be related to early aggressive blood pressure (BP) reduction. However, the relevance, etiology, and associated factors that may better allow for patient specific BP management, have not be described. We sought to identify the prevalence of this cohort at our institution. We hypothesize that aggressive blood pressure reduction in the first 24 hrs increases the likelihood of DWI abnormalities on MRI.

This study is a single center retrospective review from 2009-2012 of ICH patients who underwent MRI imaging within 10 days of ICH. Demographic, relevant past medical history, severity of illness scales, length of mechanical, ICU and hospital length of stay were collected.

To date, our analysis includes 49 and 37 patients with DWI and without DWI changes, respectively. Patient with DWI abnormalities had higher mean ICH score (2.2+1.2 vs 0.8+0.7, p<0.05), comparable age (60+13 vs 63+20, p=NS), greater BP drop (93+31vs 77+27mmHg, p = 0.02), longer mean ICU (11+7 vs 4+5, p<0.05) and hospital (14+57 vs 9+10, p=NS) lengths of stay, and higher mortality (14% vs 8%, p=0.3).

In our initial univariate analysis, DWI abnormalities were associated with greater blood pressure drop and worse outcomes. Our next endeavor is to identify the incidence and predictors of DWI changes by collecting echocardiographic and electrocardiographic measures of left ventricular hypertrophy (LVH) as a surrogate of chronic hypertension and altered cerebral autoregulation. Measuring interactions between markers of LVH, ICH size and BP reduction may better allow for BP optimization.

Aneurysmal subarachnoid hemorrhage (aSAH) is still associated with a high morbidity and mortality. A substantial amount of evidence from animal models indicates that early brain injury (EBI) may play an important role in the patient-s outcome. Cerebral microdialysis allows online measurement of brain metabolic changes and extracellular proteins including the proinflammatory cytokine interleukin-6 (IL-6) and matrix metallopeptidase-9 (MMP-9).

Twenty-six aSAH patients with multimodal neuromonitoring were analyzed in a prospective observational cohort study. Daily cerebral microdialysates were additionally analyzed for IL-6 and MMP-9 (ELISAs). Statistical analysis was performed using a generalized estimating equation with an autoregressive function to handle repeated observations within a subject.

In the first 24 hours the patient`s metabolic CNS profile revealed brain metabolic distress and an excitatory response significantly improving over the following 5 days (P<0.001). Moreover we observed an increased glucose consumption reflected by a significant decrease of brain extracellular glucose concentration (P=0.001). Brain tissue hypoxia (P bt O 2 <20mmHg) was observed in >60% of neuromonitoring time in the first 24 hours (median=15mmHg, IQR=5-22) and improved thereafter (P<0.05). Baseline IL-6 and MMP-9 levels were initially elevated (median=4059pg/ml, IQR=1316-12456; median=851pg/ml, IQR=98-25860) and significantly decreased over the next 5 days (P<0.05). A higher proinflammatory response was associated with the development of delayed cerebral ischemia (P=0.04), whereas loss of consciousness at ictus, admission disease severity and early brain tissue hypoxia were associated with higher MMP-9 levels (P<0.03). All models were adjusted for probe location, aneurysm securing method and disease severity as appropriate.

In this study we could confirm pathophysiologic mechanisms of early brain injury in patients with aSAH, reflected by a brain extracellular proinflammatory response, MMP-9 upregulation and cerebral metabolic distress. These results need to be confirmed in a larger cohort and may be used as endpoints for future interventions targeting EBI in poor-grade aSAH patients.

The pathophysiology underlying contusion expansion in traumatic brain injury (TBI) is not completely elucidated. We hypothesized that impaired autoregulation contributes to contusion expansion in TBI patients.

We retrospectively reviewed the medical records of prospectively collected TBI patients from January 2009-December 2012 at a large academic center. Patients with TBI and traumatic contusions were included if they had two head computed tomography (CT) scans within 5 days of injury and had pressure reactivity index (PRx) data available. PRx, a moving correlation coefficient between arterial blood pressure and slow waves of intracranial pressure, was calculated using ICM+ software. Contusion volume was quantified using a grid-counting method in the software image-processing package Fiji Image Just. An increase in contusion volume by ≥ 30% between scans was considered contusion expansion.

Twenty-four patients with TBI and contusions were included. Median age was 28 years and 17 (70.8%) were men. Median initial Glasgow Coma Scale was 7 (IQR 4-8). The median Marshall CT grade was 2 (IQR 2-5). The interval between CT scans varied, with median of 64 hours . Contusion expansion occurred in 19 patients (79%). Of 5 patients with stable or decreasing contusion volumes, mean PRx was <0.25 during the hour preceding the second CT in all 5, while mean PRx was ≥ 0.25 in 14/19 patients with contusion expansion. Expansion occurred in all patients (n=5) with mean 1 hour PRx ≥ 0.25. The change in contusion volume between CT scans was associated with Prx averaged over the hour preceding the second scan (p=0.049).

In this small sample of TBI patients, mean PRx was associated with contusion expansion. This suggests that impaired cerebral blood flow autoregulation may contribute to the pathogenesis of contusion expansion, but this should be considered hypothesis-generating data and deserves further study.

Financial Support: ICM+ Software is licensed by Cambridge Enterprise, Cambridge, UK. Marek Czosnyka and Peter Smielewski have a financial interest in a fraction of the licensing fee.

Quantitative EEG (qEEG) software is used to help manage large volumes of data generated by continuous EEG monitoring (cEEG) in the Neuro-ICU. However, little data is available to guide the use of qEEG in clinical practice. This study aims to evaluate the sensitivity and specificity of qEEG trends for seizure detection in adult patients in the Neuro-ICU.

QEEG panels from 45 patients admitted to the Duke Neuro-ICU (30 with and 15 without NCS) were distributed to 5 electroencephalographers and 4 EEG technicians. Each qEEG panel (n=180) consisted of Rhythmic Run Detection and Display, fast Fourier transform, amplitude integrated EEG and EEG asymmetry index. The corresponding raw EEG segments were reviewed independently to identify seizures. The reviewers did not have access to the corresponding raw EEG data.

The sensitivity was 0.85 (95% CI 0.82-0.88) and 0.80 (95% CI 0.76-0.84) with specificity of 0.66 (95% CI 0.61-0.70) and 0.73 (95% CI 0.67-0.78) respectively for electroencephalographers and EEG technicians for the ability to detect the presence of seizures on qEEG panels when compared to independent raw EEG review. The positive predictive value was 78% (95% CI 74-81%) and 81% (95% CI 77-84%) respectively for electroencephalographers and EEG technicians. Reviewers from both correctly identified the number of seizures 57% of the time. The intraclass correlation coefficient was 0.74 for all reviewers.

Commonly used qEEG trends by electroencephalographers and EEG technicians demonstrated acceptable sensitivity and specificity for the detection of the presence of seizures. Their ability to quantify the number of seizures correctly occurred more than half of the time. QEEG trends do not appear to be sufficient alone for reviewing cEEG data but may be a means for non-neurophysiologists to identify periods of concern during live recording.

Sleep in the ICU is frequently dysfunctional, and loss of sleep-like features on cEEG after acute brain injury has been associated with outcome. We hypothesized that appearance of sleep-like features on cEEG after cardiac arrest would correlate with outcome.

All cardiac arrest patients admitted to Columbia University Medical Center were prospectively enrolled into an ongoing multidisciplinary study of recovery of consciousness (ROC). Daily coma recovery scores (revised; CRS-R) were assessed during cEEG monitoring, and best sleep-like state was assigned to categories: 1=monophasic/burst-suppression; 2=alternating patterns without sleep-like features; 3=rudimentary sleep-like features; 4=normal-appearing non-REM sleep. Outcomes were survival to discharge and ROC at any time during hospitalization (CRS-R>10).

14 patients were enrolled. Mean age was 69+/-12; 79%(n=11) were male. Initial rhythms were PEA(n=7); VF(n=4); and asystole(n=3). 43%(n=6) were in-hospital; all were witnessed, and 79%(n=11) received bystander CPR. Median time to return of spontaneous circulation (ROSC) was 11.5 min ; median initial GCS was 3 (IQR 3-3). 43%(n=6) survived to discharge; of those who died, all but one had care withdrawn. 36%(n=5) recovered consciousness, with a median best CRS-R of 17 (IQR 14-23). Age, ROSC, initial rhythm, lactate, initial neuron-specific enolase, and seizures on cEEG did not associate with ROC or survival. The presence of sleep-like features was associated with survival (p=0.04) and ROC (p<0.01). Among those with ROC, the best sleep-like features were seen prior to the best CRS-R (1.4 vs 2.5 days post arrest; p=0.07). Reactivity on cEEG was also associated with survival (p<0.01) and ROC (p=0.01) .

This prospective study demonstrated that appearance of sleep-like features on cEEG was associated with both ROC and survival, and suggested recovery of sleep-like features may precede maximal ROC. This, along with other studies, points to sleep as a prognostic marker after acute brain injury, particularly cardiac arrest.

Spontaneous intracerebral hemorrhage (sICH) location is often considered singular although hemorrhages do not respect anatomic borders. We hypothesized that number of anatomic subregions involved in sICH decreases with time and is associated with long-term neurologic outcomes.

This was a post hoc exploratory analysis of computed tomography (CT) brain scans collected from three trials evaluating acute therapies for sICH (CLEAR IVH, ICES, MISTIE II). Anatomic subregions involved were recorded for CT images taken at ICH onset, 30 days, and 180 days. Defined subregions included lobar (frontal, parietal, temporal, occipital), basal ganglia (caudate, putamen, globus pallidus), internal capsule, and thalamus (anterolateral, posterolateral, medial, dorsal) .

Of 170 patients, primary ICH location was identified as lobar in 42 (24.7%), basal ganglia in 106 (62.4%), and thalamus in 22 (12.4%). Number of anatomic subregions involved on both diagnostic CT and day 30 CT were significantly associated with day 30 poor outcome (mRS 4-6 vs. mRS 0-3) (diagnostic: 3.4±1.5 vs. 2.6±1.2, p=0.03)(day 30: 3.7±1.6 vs. 2.4±1.2, p=0.01). Greater number of regions at day 180 only was associated with poor outcome at day 180 (4.0±1.8 vs. 2.9±1.6; p=0.01). After adjustment for ICH score, number of regions remained significantly associated with poor outcome (day 30: p=0.003; day 180: p=0.05). Number of regions involved increased between diagnosis and day 30/day 180 CT in patients with poor outcomes and decreased in patients with good outcomes (day 30: increase of 0.5±1.4 vs. decrease of 0.2±0.9 regions, p=0.04; day 180: increase of 0.7±1.4 vs. 0.0±1.1 regions, p=0.04).

Number of anatomic subregions involved at days 30 and 180 are significantly associated with neurologic outcome independent of ICH score. Evolution of ICH lesions over time may play a role in recovery. Mechanism of this observed late increase in anatomic distribution represents a topic of further research.

Financial Support: Dr. Hanley has received support from NINDS for CLEAR III 5U01-NS062851-03 and MISITE II 5R01-NS046309-07

Estrone (E1) and Estradiol (E2) are hormones with diverse actions on the neurovascular unit and central nervous system however data on the role of endogenous estrogens in acute aneurysmal subarachnoid hemorrhage (aSAH) are sparse. Our aim was to investigate the role of circulating estrogen levels on outcomes in acute aSAH patients.

Plasma samples were collected on 99 acute, adult aSAH patients admitted to the NV-ICU enrolled in a NIH funded study (RO1NR004339). Three plasma samples were selected for estrogen analysis from each patient representing early (1) (2) (3) (4) , middle(4-6) and late(7-10) days after hemorrhage and were assayed using liquid chromatography-tandem mass spectrometry. Functional outcome and mortality was evaluated at 3 and 12-months after hemorrhage by Modified Rankin Scale (MRS) [dichotomized as good (1) (2) (3) and poor(4-6)]. Statistical analysis using SPSSv19 and SASv9.2 included correlations, group based trajectory and multiple logistic regression.

Higher E2 levels were significantly associated with death (β=.245,p=.003) and outcome at 3-months after hemorrhage. Controlling for age and admission Hunt and Hess(HH), higher mean E2 concentrations remained significantly associated with poor MRS at 3-months [OR 2.8, 1.3-6.2(95% CI),p=.01]. Controlling for age and HH, higher mean E1 concentrations also remained significantly associated with poor MRS at 3-months [OR 2.6,1.3-3.6(95% CI),p=.01]. Group based trajectory identified 2 distinct populations over time for both E1(61.4% high,38.6% low) and E2(48% high,52% low) values using censored normal model. Patients in the low E2 group were significantly more likely to have good 3-month(p=.004) and 12-month(p=.02) outcomes. Likewise patients in the low E1 group were significantly more likely to have good 3-month outcomes by MRS(p=.01).

These results provide the first clinical evidence of an independent association between higher endogenous estrogen levels and mortality and poor outcomes and provide incentive for future studies to further investigate the role of estrogen in patients with aSAH.

Financial Support: Funding for this project was supported by NIH, the Neuroscience Nursing Foundation and the Nightingale Scholarship of Pennsylvania Introduction ICH Score is a well-validated instrument that evaluates age, ICH location and size, and presence of IVH to predict 30-day mortality. Potentially modifiable factors that further increase risk of mortality during hospitalization after ICH remain incompletely understood, and might serve as useful targets for interventions.

From September 2009 to December 2012, 286 patients with ICH admitted to the neurological ICU at Columbia University Medical Center were prospectively enrolled in the ICH Outcomes Project (ICHOP). In-hospital mortality was selected as the primary outcome.

Mean age was 62.2 +/-18.5 years; average APACHE-II score was 15 +/-8.6 (predicted mortality 8-25%); and average ICH score was 1.9 +/-1.3 (predicted mortality 13-39% at 30 days). Of the 286 patients 73 (26%) died; of those who died, 93% (68/73) had a DNR order at the time of death. In a bivariate logistic regression model the following factors were predictive of in-hospital mortality: ICH score, secondary ICH due a vascular anomaly (protective), any fever >38.5 °C, pneumonia, and respiratory failure requiring mechanical ventilation. Using the Delong Delong method for comparing ROCs; AUC for the model with ICH score was 0.87 while the AUC for the model without ICH score, was 0.95. After accounting for ICH score, these additional factors explained acute mortality more accurately and improved the overall predictive value of the model. Despite associations with mortality in univariate analysis, the following factors were not associated with mortality after accounting for other variables in the final model: APACHE-II score, pre-morbid modified Rankin score, surgical evacuation, EVD placement, cocaine use, neurological deterioration, anemia, and sepsis.

After accounting for admission non-modifiable predictors of mortality accounted for by the ICH score; fever, pneumonia, and respiratory failure significantly increase mortality risk. Innovative strategies focused on fever control, respiratory support, and prevention of hospital-acquired pneumonia hold promise to reduce mortality after ICH.

We investigated combinations of three components of the Glasgow Coma Scale (GCS); eye opening (E), verbal response (V), and best motor response (M) in acute impaired consciousness associated with disease.

We identified each combination of E, V, and M level of 3540 patients transported to our medical center with impaired consciousness, selected from 7645 patients by excluding cases with cardiac arrest, younger than 6 years, or systolic blood pressure lower than 90mmHg. We categorized patients into 6 groups: (1) traumatic brain injury (total number 760), (2) cerebral stroke (566), (3) acute intoxication (718), (4) seizure (290) , (5) metabolic disease (246), and (6) other disease (1187). We calculated odds of E1-4, V1-5, and M1-6 in each group using logistic regression analysis.

We obtained odds ratios for all levels of the three components for each group by comparing odds to the lowest one (odds: 1) in each component. Next, we made a multiplication table consisting of 120 values, which means all combinations of GCS, for each group. Each value in the table was calculated by multiplication of three odds ratios with three components.We found some combinations that showed an extremely high product and a strong propensity to the combinations were observed in the three groups: E1V1M2 in stroke, E2V3M5 in acute intoxication, and E4V1M1 in seizure.

A high propensity identified in stroke, acute intoxication, and seizure may indicate the existence of different mechanisms of impaired consciousness in these diseases. GCS may have a potential to represent these mechanisms.

Underlying etiologies of encephalitis often remain unknown despite extensive workup. Cognitive and epilepsy-related outcomes are poorly studied, but early diagnosis and treatment may improve outcome.

We prospectively screened all patients with presumed encephalitis from July 2012-January 2013 and applied a comprehensive diagnostic workup. Diagnosis was confirmed with biopsy or ≥2 biologically plausible and concordant serum, CSF, or other tests; diagnosis was presumed if testing was positive but not concordant. Observational data was collected along with functional and cognitive outcomes measured at 6-12 months.

We enrolled 14 patients. Median age was 49 (IQR 38-66); 29% (n=4) were women. Clinical seizures were documented in 10/14; 6/14 had refractory status epilepticus (RSE). Median transfer time after symptom onset was 10 days (IQR 5-12). Workup included serum encephalitis screening (14%); serum paraneoplastic panel (64%); CSF encephalitis screening (64%); CSF paraneoplastic panel (64%); and CSF HSV PCR (93%). 3 patients had brain biopsy and 1 had autopsy. Only 1/14 had confirmed diagnosis; 4/14 had presumptive diagnosis; 9/14 remained idiopathic. 7/14 received immunotherapy; 4/6 with RSE received immunotherapy. Immunotherapy was started median 13 days (IQR 8-14) from symptom onset and 1 day (IQR 1-3) from admission. Overall, 4/14 (2 with RSE) died at hospital discharge. Follow-up data at median 7 months (IQR 6-7.8) for 8/10 surviving patients showed median modified Rankin score 1 (IQR 0-3) and median Telephone Interview for Cognitive Status 35 (IQR 29-40). 1/8 had positive epilepsy screen, 7/8 were on anticonvulsants, and 6/8 were seeing an epileptologist.

Despite a comprehensive, standardized diagnostic workup for patients presenting with encephalitis, underlying etiology remained idiopathic in 64%. Although discharge mortality was 29%, survivors had normal cognition despite non-impairing deficits after 6 months. We are currently running MassTag PCR on CSF, serum, and feces to detect novel viral causes in all 14 patients.

Therapeutic plasma exchange (TPE) is a well-established treatment for various neurological disorders but in cases of encephalitis with suspected immune etiology or cases refractory to immunosuppressive therapy, its use is yet to be defined.

A retrospective study of acute encephalitis in ICU who received TPE at the Johns Hopkins Hospital (JHH) from July 2001 to June 2013. The patients were divided into an established indication (EI) for TPE group by American Society of Apheresis guidelines and non-established indication (NEI) group.

Of the 31 patients included, 5 were in the EI group (M=4, F=1); and 26 (M=11, F=15) in the NEI group. Mean age was 41.7±19.9 (5-80). EI group included 4 ADEM cases and 1 NMO case; NEI group included 12 patients with antibodydetected encephalitis and 14 with non-antibody detected encephalitis. At the time of first TPE treatment: GCS (EI 11±4; NEI 10±4; GCS £8 (EI 20%; NEI 35%); Intubated (EI 40%; NEI 50%); Status Epilepticus (EI 0%; NEI 31%); Length of stay (los) (EI 24±2 days; NEI 40±3 days). Most patients received 5 to 7 TPE treatments. All patients survived with overall incidence of adverse events (AE) of 2.61% per procedure (5/191 procedures) . Five AE were reported in the NEI group: hypotension, catheter infection (2), allergic reaction and UGIB. Modified Rankin Scale at discharge was similar for both (EI (3.8±1.1) and NEI (3.1±1.9)) groups; p=0.4.

TPE appears to be tolerated well in this critically ill population, with no deaths and comparable functional outcome between groups studied. The higher AE in the NIE group was expected and manageable, and likely related to poorer clinical status at the time of TPE. Future studies are needed to better define the role and efficacy of TPE in this challenging population.

Intraventricular administration of antimicrobials (IVRabx) may be necessary in certain difficult-to-eradicate central nervous system infections where morbidity and mortality are high. Despite increasing prevalence of such infections, there are no guidelines for IVRabx management and current practice is poorly understood.

Retrospective review of patients admitted to a Neurosciences ICU from 2005-2013 who received IVRabx. Clinical and laboratory data was collected. Survivors and non-survivors were compared.

Of 22 patients (M=17;F=5), there were 14 survivors (mortality 36%). Mean age was 51.5±3.9 (SEM) years. The most common admission diagnoses were ventriculoperitoneal shunt infection (N=6) and brain tumor (N=8). Both meningitis and ventriculitis were present in 18%; 32% of infections were associated with external ventricular drain use. Cerebrospinal fluid (CSF) profile at diagnosis differed between survivors and non-survivors: median WBC (survivors: 504(iqr 1191)/cu mm; non-survivors: 3288(8042); p=0.04); glucose (survivors: 46(47)mg/dl; non-survivors: 2(7.5); p=0.005). In-hospital mortality was non-significantly higher for gram-negative vs. gram-positive infections (50% vs. 22.1%; p=0.37). The most common isolated organisms were Staphylococcus aureus (28.6%) and Enterobacter cloacae (23.8%). All patients received treatment with both systemic and IVR antibiotics. IVRabx were: Vancomycin(7), Gentamicin(4), Tobramycin(6), Vancomycin+Gentamicin(2), Gentamicin+Tobramycin (1) and Amikacin(2). Therapeutic CSF drug monitoring was obtained in 71%(survivors) and 88%(non-survivors) of cases respectively (p=0.61). Median days from infection diagnosis to starting IVRabx was non-significantly longer in survivors (6(9) vs. 2.5(3)-non-survivors; p=0.08). Median days to CSF sterilization was non-significantly longer in survivors (8.5(17) vs. 5.5(5.5)-non-survivors; p=0.16). Three patients (13.6%) had recurrent CSF infection.

Addition of IVRabx to intravenous treatment was started early and with high rates of therapeutic monitoring. In-hospital mortality was high despite shorter times to start IVRabx and to sterilize CSF compared to survivors. Severity of initial infection was significantly associated with mortality. Further studies are needed to characterize criteria, optimal dosing regimens and patient specific factors which impact efficacy of this approach.

Current methods to evaluate cerebral edema are either invasive (ICP monitoring) or intermittent (neuroimaging). The ability to non-invasively measure small changes in intracranial fluids could potentially identify edema earlier. The purpose of this study was to assess feasibility of volumetric integral phase-shift spectroscopy (VIPS) technology to non-invasively assess intracranial fluid content. Additionally, the role of the neurocritical care unit in device development clinical research is also described.

The VIPS monitor involves placement of a sensor on each side of the head so that wave frequencies between 10 and 500 MHz pass through the frontal lobes. Pre-clinical studies have shown that shift in phase angle relates to tissue and fluid composition. Subjects were eligible if they had an acute brain injury and were undergoing ICP monitoring or serial neuroimaging. As this was a feasibility study, aspects related to VIPS device configuration were also evaluated and changed to improve device performance. Human subjects approval was obtained.

Informed consent for participation was obtained for 22 subjects. Usable VIPS data was acquired from 11 subjects. Initial monitoring was undertaken for 2-4 hours in 6 subjects to test feasibility, with overnight monitoring then undertaken. ICP values were compared to VIPS values for individual patients through visual inspection of graphs which superimposed the two parameters after normalization of the VIPS baseline value. In 9 subjects, at least a portion of the ICP tracing tracked with the VIPS tracing. In one patient with global cerebral ischemia after cardiac arrest, a dramatic increase in the VIPS value preceded the clinical event of central herniation. Device configuration was modified twice in order to improve signal quality and limit artifact.

VIPS cerebral monitoring is feasible and may correlate with ICP and cerebral edema. The neurocritical care unit is a useful clinical research environment for early stage monitoring device development.

Financial Support: This research is supported by the developer of the VIPS device, Cerebrotech.

In shock, the venous-arterial carbon dioxide difference may be an indicator of tissue perfusion, with increased values suggesting ischemia. 1, 2, 3 By using a jugular bulb source, the venous-arterial carbon dioxide gradient for the brain can be calculated (v-a(CO2) cer , cer=cerebral). V-a(CO2) cer increases as patients progress to brain death, 4 but no previous study has examined this metric after brain injury or correlated it with outcome.

This prospective observational study investigated patients with brain injury. Brain tissue oxygenation was measured with a parenchymal probe; local cerebral blood flow was measured with a parenchymal thermal diffusion probe. Cerebral venous blood gases were drawn from a jugular bulb venous catheter. The data were analyzed for v-a(CO2) cer differences, surgical interventions, probe placement, cerebral blood flow, brain tissue oxygenation and patient outcome. A good outcome was defined as discharge to home or a rehabilitation facility; poor outcome was defined as death or discharge to a long-term care facility.

Of 21 patients with traumatic brain injury (n=17), stroke (n=3) or intracerebral hemorrhage (n=1), twelve patients had a good outcome and nine patients had a poor outcome. Probe placement, surgical interventions, and the Glasgow Coma Scale (GCS) in the emergency room were similar across the groups. In patients with a good outcome, the mean va(CO2) cer was 9.00 (+/-6.57) mmHg; in patients with a poor outcome mean v-a(CO2) cer was 5.75 (+/-4.00) mmHg, p<0.03. Cerebral blood flow differed significantly as well (17 +/-13 cc/100gm/minute vs. 29 +/-14 cc/100gm/minute, p<0.05, good vs. poor outcomes). Brain tissue oxygenation did not differ between the groups.

Elevations in v-a(CO2) cer correlated with a good neurologic outcome. Patients with good outcomes also had significantly lower cerebral blood flow. Brain tissue oxygenation did not differ significantly. Relative hypoperfusion may be beneficial in the initial period after injury, particularly in patients with impaired cerebral autoregulation, though additional research is needed.

To assess the causes, management, and outcome of patients with presumed autoimmune or viral encephalitis.

All patients admitted between 2008-2011 with spinal fluid (CSF) sample sent for herpes PCR, paraneoplastic panel, or New York State encephalitis panel were retrospectively reviewed. Patients fitting case definition (new onset encephalopathy with clinical, spinal fluid, EEG, or MRI evidence of inflammation) were included. Etiologies were categorized based on clinical scenario; treatment and outcome at hospital discharge were compared.

114 patients met case definition of encephalitis. Of those, 40 had alternative causes: toxic/metabolic(n=10), bacterial/fungal(n=9), neoplastic(n=9), neurodegenerative(n=6), parasitic(n=3), new-onset epilepsy(n=2), acute psychosis(n=1). Excluding these, 65%(n=74) had presumed autoimmune or viral encephalitis; mean age 50.25+/-18.6; 42%(n=48) were women. Workup revealed viral etiology in 11%(n=8); autoimmune etiology in 26%(n=19); 64%(n=47) remained idiopathic. 36%(n=17) had RSE, 23%(n=11) had seizures/SE, and 40%(n=19) had no seizures. Immunotherapy was used in 31%(n=35) of the whole cohort at mean 48+/-61 days from symptom onset. Discharge mRS was not different between those who received immunotherapy vs those who did not. Immunotherapy was used in 23%(n=11) of idiopathic cases at mean 26+/-27 days from symptom onset and 35%(n=6) of RSE cases at mean 16+/-8 days from symptom onset; neither showed difference in discharge mRS. Mortality was 16%(n=18) overall, 21%(n=10) in idiopathic cases and 53%(n=9) in those with RSE. Half of idiopathic cases had good outcome compared to only 12%(n=2) of RSE cases (p<0.01).

Most patients with presumed autoimmune or viral encephalitis remained without proven cause after diagnostic workup. The use of immunotherapy was not associated with better outcome, but was on average given more than two weeks from symptom onset. Better diagnostic strategies and perhaps earlier treatment might be targets to improve outcome in idiopathic cases.

Management of intracranial pressure (ICP) utilizes monitors, which provide a single value for the cranial vault irrespective of lesion location and size. Prevailing practice assumes that increased ICP is the dominant cause of brain herniation and poor outcomes. In this study we hypothesize that the occurrence of clinical brain herniation does not correlate with global ICP elevation.

A prospective observational cohort study of acutely comatose patients admitted to the NCCU was performed between May 2010 and May 2011. Brain codes were defined as new onset of coma with pupillary change. Clinical brain herniation (CBH) was defined as acute onset of unilateral or bilateral pupillary dilation with loss of reactivity to light, and a decline of GCS≥2.

142 patients were enrolled in the study, of which 68 had a brain code. 54 suffered from CBH and 14 from a herniation mimic. Etiology of CBH were: SAH(42%), ICH(26%), TBI(17%), ischemic stroke(7%), tumor(6%). CBH was transtentorial(63%), subfalcine(33%), tonsillar(1%). 44 CBH were due to a supratentorial lesion and 10 to an infratentorial lesion. 55%(30/54) had ICP monitors (87%EVD, 10%codman, 3%bolt), that were placed after the initiation of the neurological deterioration. In patients with brain herniation from an infratentorial lesion and in 22% of patients with a supratentorial lesion the ICP was normal. All patients who had CHB had poor outcomes (mRS >2) at discharge, irrespective of normal or elevated ICP.

ICP was not elevated in all patients with CBH supporting the concept of compartmentalized ICP. The timing and site of the ICP monitor as well as lesion character and therapies provided may limit our observations. Management of patients with CBH should not neglect clinical examination in favor of ICP monitoring. There remains the need to improve monitoring, which reflects not only singular ICP, but also compartmental tissue pressures, where brain injury is immediately occurring.

The majority of comatose survivors of cardiac arrest have poor long-term functional outcomes; though the use of therapeutic hypothermia has improved the number of survivors who go on to have a good outcome. Little is known about the ideal physiologic targets such as blood pressure and oxygenation in the acute period after cardiac arrest. One animal study showed improved outcomes with blood pressure augmentation, and early goal directed hemodynamic optimization has been suggested in humans in the post-arrest period.

This pilot study is based on data from a prospective observational trial investigating comatose survivors of cardiac arrest. Hourly hemodynamic data for the first 72 hours after the arrest were recorded. Patients were divided into good and poor outcome groups based on a modified Rankin score at 6 months of 0-3 vs. 4-6, respectively. Association of mean arterial pressure (MAP) with clinical outcomes was assessed.

23 initial consecutive patients were included in this analysis: 15 patients had a poor outcome and 8 patients had a good outcome. Of those with poor outcome, 13 died and 2 survived with significant disability. The MAP within the 48-72 hour post-arrest period was significantly higher in those who survived with good outcome (median [IQR] 

Higher mean arterial blood pressure in the acute period after cardiac arrest correlated with improved long-term functional neurologic outcomes. Additional analysis is planned to investigate a larger dataset and to compare the blood pressure with magnetic resonance imaging parameters. In the acute period after cardiac arrest, higher blood pressure may be beneficial to long-term neurologic outcome.

Steroid-responsive encephalopathy with associated autoimmune thyroiditis (SREAT, "Hashimoto encephalopathy") is diagnosed in patients with steroid-responsive encephalopathy in which infectious and metabolic causes have been exluded and elevated serum TPO antibodies are present. The goal of the present study was to determine in a single institution setting whether patients first diagnosed with SREAT were given a revised diagnosis at follow-up.

We reviewed charts of 405 patients admitted to the department of neurology during a 10 year period, in whom the ICD-10codes of encephalopathy of unknown origin or autoimmune thyroiditis were given. In patients in whom the diagnostic criteria were fulfilled and who were diagnosed with SREAT basic clinical data, presence of serum antibodies (apart from TPO-antibodies), presence of cranial MRI changes and outcome (as documented at consecutive admissions) were assessed.

In 10 Patients SREAT was diagnosed (9 females). Follow up information was available in 8 patients. In 2/8 patients Anti-NMDA-receptor (NMDAR) encephalitis and in 1/8 patient Adult Onset Still's disease (AOSD) were diagnosed. Outcome was favorable in 6 patients, one patient had neurologic deficits and one patient (with AOSD) died

In more than one third of patients in whom follow up data were available (3/8) the diagnosis of SREAT had to be corrected. NMDAR-encephalitis and rare rheumatologic conditions, e.g. AOSD, should always be considered in patients who fulfill the diagnostic criteria of SREAT. 

Prognostication after traumatic brain injury (TBI) is critical in the neurological intensive care unit (ICU). Accurate shortand long-term outcome predictions in TBI are driven by the need of affected families for decision-making. In clinical practice, it is unknown whether clinicians actually apply the IMPACT prediction model and how clinicians from different specialties prognosticate outcomes. In this study, we examined the variability of TBI outcome prognostication among members of the Neurocritical Care Society (NCS).

We conducted a national web-based survey of all members of the NCS over a two-week period in August 2012.

143 NCS members participated in the survey with years of experience ranging between 6 to 15. Seventy-four percent (106) provide TBI outcome prognostication. Sixty-two percent (88) of our cohort were trained in neurocritical care. Interestingly only 18% (17) of participants use the IMPACT model for prognostication while 42 % (40) are not aware of IMPACT predictors. Seventy-six percent (72) of participants make their prognostication based on physical exam/radiological variables, with some influence by "Gestalt" and "experience". Fifty-eight percent (55) of participants think that physicians from different specialties prognosticate differently. Pupillary reactivity,age, Glasgow coma scale and pre-admission hypoxia are the prognostic variables considered most important to outcomes.

Our survey supports our hypothesis that within the NCS community wide-ranging variability in TBI outcome predictions exist. This might be driven by the provider experience level and educational background. IMPACT predictors are not widely used by survey respondents. Efforts should be undertaken to standardize prognostication in TBI. One solution may be a shared-decision aid in TBI.

Traumatic brain injury (TBI) is associated with loss of cerebrovascular autoregulation, which leads to cerebral hypoperfusion. Mitogen activated protein kinase (MAPK) isoforms ERK, p38, and JNK and endothelin-1 (ET-1) are mediators of impaired cerebral hemodynamics after TBI. Excessive tPA released after TBI may cause loss of cerebrovascular autoregulation either by over-activating N-methyl-D-aspartate receptors (NMDA-Rs) or by predisposing to intracranial hemorrhage. Our recent work shows that a catalytically inactive tPA variant (tPA-S 481 A) that competes with endogenous wild type (wt) tPA for binding to NMDA-R through its receptor docking site but that cannot activate it, prevents activation of ERK by wt tPA and impairment of autoregulation when administered 30 min after FPI. We investigated the ability of variants that lack proteolytic activity but bind/block activation of NMDA-Rs by wt tPA (tPA-S 481 A), don't bind/block activation of NMDA-Rs but are proteolytic (tPA-A 296-299 ), or neither bind/block NMDA-Rs nor are proteolytic (tPA-A 296-299 S 481 A) to prevent impairment of autoregulation after TBI and the role of MAPK and ET-1 in such effects.

Lateral fluid percussion brain injury (2 atm) was induced in anesthetized piglets equipped with a closed cranial window. Cerebral blood flow was determined through use of neutron activated microspheres. Autoregulation was investigated by inducing hypotension through withdrawal of blood to decrease mean arterial pressure (MAP) by 25% (moderate) or 45% (severe). Such drops in MAP were held constant for 10 min via titration of blood withdrawal/shed blood re-infusion. Pial artery diameter and CBF were determined during normotension and hypotension. CSF concentration of p38 MAPK and ET-1 was determined by ELISA.

Non-invasive cranial accelerometry detects minute vibrations in the skull produced by pulsatile cerebral blood flow. We explored the possibility that athletes suffering mild concussion would have changes in cerebral hemodynamics measureable with cranial accelerometry.

Cranial accelerometry was performed using a prototype device designed by Jan Medical, Inc. (Mountain View, CA). Subjects were high school football players who consented for baseline recording and multiple recordings following a reported suspected concussion event (SCE). SCAT2 was documented at baseline and during the concussive phase. Frequency analysis (Fast-Fourier transform) of accelerometry recordings was performed on all recorded epochs and subjected to visual and numerical analysis of power spectra.

All subjects at baseline had typical accelerometry recordings showing 3 or 4 harmonic peaks of the baseline cardiac cycle. However, subjects who reported an SCE typically had a shift of vibration frequencies at 5 to 8 harmonic peaks of the cardiac cycle. Using this dichotomy, of the 91 non-SCE subject recordings, 79 recordings did not show these higher harmonics but 12 did (specificity 87%, false positive rate 13%). Of the 17 SCE subjects, 13 (16% of subjects) were felt clinically to have sustained a concussion based on SCAT2 and the judgment of the coach/team physician. Of these 13, accelerometry detected a shift to higher frequency in 10 (sensitivity 77%) but not in 3 (false negative 23%).

Highly sensitive cranial accelerometry detects an apparent shift toward higher harmonics of the cardiac cycle in patients with concussion. The cause of this shift may be related to regional loss of cerebral autoregulation or cerebral edema. This non-invasive technique may be useful to objectively document the presence of a concussion and to determine the time to resolution.

Financial Support: WSS has received grant support from Jan Medical, Inc.

Pediatric head injury is a common ED presentation.In contrast to adults,brain imaging orders are less due to risk of radiation to developing brain.We sought to determine if any clinical presentation features were significantly associated with having an abnormal head CT,or neurosurgical intervention.

This IRB approved study is an observational cohort of children aged 4-17 years who presented to our pediatric emergency department. 

The majority of patients were presented within 12 hrs(86%),7% presented between 12-24 hrs,and 7% 24 hrs after injury.Univariate least squares analysis revealed that TLOC(p<0.0001);AOC(p=0.0018);and PTA(p<0.0001) were significant predictors of abnormal head CT.In multivariate model,PTA was statistically significant predictor(p=0.016).For the outcome neurosurgical intervention,all variables were significant in the univariate model.In multivariate regression model,only AOC(p=0.05) and PTA(p=0.003) retained statistical significance.

Our pilot data suggest that children presenting with PTA or AOC individually or togetherhad more probability of having a bleed or fracture on head CT along with higher rates of neurosurgical intervention.These data may help to risk stratify children with regard to need for emergency brain imaging.

We aim to investigate whether Therapeutic Induced Normothermia (TIN) ≤ 37.5°C, by means of intravascular cooling devices is more efficacious than standard medical therapy (MED) in alleviating metabolic crisis (MC) acutely following Traumatic Brain Injury(TBI).

We retrospectively analyzed data from 62 patients with severe TBI, GCS ≤ 8. We divided the cohort into two groups.1) Patients who had temperature controlled via standard medical therapies (n=52), 2) TIN group (n=10). For each group we analyzed the percent time spent in normothermia, and in MC. Additionally, for the TIN group we compared temp, ICP, sedation and MC before and after intravascular cooling. Outcomes data are underway. 

The principal results are that therapeutic normothermia using intravascular cooling results in a reduction in the burden of metabolic crisis acutely following TBI as measured by cerebral microdialysis. This differential effect occurs despite equivalent control of ICP in both TIN and MED treatments. These results demonstrate proof of concept of normothermia being neuroprotective.

The effects of antiplatelet agents on coagulation pathways in traumatic brain injury are not well understood, but available data suggest that their use increases the risk of an unfavorable outcome. We sought to determine whether pre-injury use of antiplatelet agents was associated with progression of hemorrhagic injury (PHI) or mortality after severe TBI.

Patients with severe TBI who underwent intracranial pressure and brain oxygen monitoring between 4/2008 and 12/2009 were retrospectively identified from the Brain Oxygen Monitoring and Outcome study at a University-affiliated, level-1 trauma center. Exclusion criteria included fixed and dilated pupils on admission, less then 24 hours of intracranial monitoring, or examination consistent with brain death within 48 hours of enrolment. PHI was defined as an increase in width of a subdural hematoma or volume of an intra-cerebral hematoma of >33% on repeat imaging within 24 hours from the admission scan.

Seventy-three patients were included with a median age of 45.8 years (range 15.5 -90.7). Mean GCS was 6.6 (SD 4.5), mean APACHE II score was 22.9 (SD 5.4) and mean Injury Severity Score was 31.5 (SD 9.5). Eighteen (24.7%) patients died. Nineteen patients (14.3%) were taking antiplatelet agents pre-injury. In univariate analysis, pre-injury antiplatelet use was not associated with PHI, but was significantly associated with mortality (p<0.001). In stepwise multivariable logistic regression analysis, only age [OR 1.07 (95% CI 1.02, 1.11); p=0.004] and platelet transfusion [OR 8.3 (95% CI 1.6, 43.0); p=0.01] remained associated with mortality.

The data suggest that premorbid antiplatelet use is not associated with PHI after TBI. However, there is an association with mortality that may result from potential deleterious effects of platelet transfusion. Whether platelet transfusion is simply a marker of disease severity, premorbid condition or an independent factor will need further study.

Introduction 'Pressure times time dose' (PTD) of intracranial hypertension (ICP>19 mmHg) is instrumental in predicting outcome after severe traumatic brain injury. We used automated data collection to measure the effect of common medications on the duration and dose of intracranial hypertension.

Patients >17 years old, admitted and requiring ICP monitoring between 2008-2010 at a single large urban tertiary care facility were retrospectively enrolled.Timing and dose of ICP-directed therapy were recorded from paper and electronic medical records. ICP data was collected automatically at 6-second intervals and processed. Percent time of intracranial hypertension (PTI) and pressure times time dose (PTD; mmHg*hr) were calculated.

98patients with 664 treatment instances were identified. Baseline PTD ranged from 27 (before administration of propofol and fentanyl) to 150 mmHg*hr (before mannitol). A 'small' dose of hypertonic saline (HTS; ≤250 ml 3%) reduced PTD by 38% in the first hour and 37% in the second hour, and reduced the time with ICP>19 by 38 and 39% after 1 and 2 hours, respectively. A 'large' dose of HTS reduced PTD by 40% in the first hour and 63% in the second (PTI reduction of 36 and 50%, respectively). Mannitol reduced PTD by 52% in the first hour and 44% in the second (PTI reduction of 63 and 45%). An increased dose of propofol or fentanyl infusion failed to decrease PTD, but reduced PTI between 14% (propofol alone) and 30% (combined increase in propofol and fentanyl, after 2 hours). Barbiturates failed to decrease PTD, but decreased PTI by 30% up to 2 hours after administration. Reductions significantly changed from baseline, p<0.05.

A wide range of baseline PTD values before drug administration reflects varied patient criticality. Hyperosmolar treatment reduced PTD and PTI burden significantly more than escalation of sedation or pain management, and this effect remained significant at 2 hours after administration.

In adult patients presenting acutely with mild traumatic brain injury (TBI), use of anticoagulants or anti-platelets prior to head injury has been reported to significantly increase in-hospital mortality. Select patients receive anticoagulant or antiplatelet reversal therapy upon arrival to the ED to reduce risk of bleeding, but the effectiveness of reversal therapy at reducing in-hospital mortality has not been reported. We aimed to assess anticoagulant and antiplatelet reversal therapy's efficacy in reducing in-hospital mortality in adult patients presenting to the ED with mild TBI.

This is an observational cohort study of adult patients who came to a Level I trauma center in North Central Florida with mild TBI. Independent variables included pre-ED anticoagulant or antiplatelet use and in-hospital reversal therapy; these were tested for correlation with in-hospital mortality. Data was entered into RedCap. Statistical analyses were performed in JMP 10.

Of the total mild TBI cohort (n=2243), 199 patients were taking anticoagulants or anti-platelets prior to head injury, and 17 of these patients received reversal therapy. Patients who received reversal therapy had a mean INR of 3.65 and mean APTT of 48.1, compared to a mean INR of 1.39 and mean APTT of 30.5 for patients who did not receive the therapy. Administration of reversal therapy was associated with significantly higher in-hospital mortality (p<0.0001).

Patients who receive anticoagulant or antiplatelet reversal therapy present with lower clotting tendency on average than patients not selected for reversal therapy, suggesting suboptimal medication management in individuals requiring reversal therapy. Following administration of reversal therapy, clotting tendency fails to significantly increase, so these patients remain at heightened risk for-life threatening bleed.

In recent years, researchers have recognized alcohol's ability to act as a blood thinner, even in moderate consumption.

Since pharmacological anticoagulants, such as Warfarin, significantly increase a patient's risk of bleed on head CT following mild traumatic brain injury (TBI), it seems plausible that alcohol would have a similar effect. The aim of this study is to determine whether alcohol's anticoagulant activity increases an adult patient's risk for bleed on head CT following mild TBI.

This is an observational cohort study of adult patients who presented to a Level I trauma center mild TBI, defined as a Glasgow Coma Scale (GCS) of 13-15 . Independent variables included GCS score and ED blood alcohol levels; these were tested for correlation with bleed on head CT. Data was entered into RedCap. Statistical analyses were performed in JMP 10.

Of the total mild TBI cohort (n=2243), blood alcohol concentration was measured in 293 patients, and 75.1% had positive alcohol levels, defined as >30 mg/dl. In these patients, alcohol consumption prior to head injury failed to predict bleed on head CT (p=0.2173). Additionally, excessive alcohol consumption, defined as ≥80 mg/dl, was not significantly associated with bleed on head CT following mild TBI (p=0.2950).

Although alcohol acts as an anticoagulant at moderate levels, alcohol consumption prior to TBI does not significantly increase a patient's risk for bleed on head CT, suggesting that alcohol's anticoagulant ability is significantly weaker than that of pharmacological anticoagulants and that alcohol's blood thinner effect may be clinically insignificant in cases of mild TBI. Therefore, it may not be necessary for ED physicians to allow alcohol consumption alone to guide their decision to obtain a head CT following mild head injury.

The ED lacks continuity of care for mild traumatic brain injury (mTBI) patients, with many never receiving a head CT following injury; however, previous research from our lab has shown that patients presenting acutely with mTBI have significantly higher rates of abnormal head CT scans if they have elevated ED glucose, defined as >130 mg/dl, or elevated WBC counts, defined as >14,000 cu/mm (p<0.0001 and p=0.0464, respectively). The current study aims to characterize these abnormalities to guide ED physicians in their decision to obtain CT scans.

This is an observational cohort study of adult patients who came to the emergency department of a Level I trauma center in North Central Florida with mTBI. Independent variables included ED glucose and WBC count; these were tested for correlation with specific abnormal head CT findings, the dependent variables. Data was entered into RedCap, and statistical analyses were performed in JMP 10.

Of the total mTBI cohort (n=2243), 286 patients had elevated ED glucose and 203 patients had elevated WBC count. Elevated ED glucose and WBC count were predictive of parenchymal or hemorrhagic contusion (p=0.0028 and p=0.0279, respectively) on head CT. Increased ED glucose alone was significantly associated with subdural hematoma (p=0.0451) and subarachnoid hemorrhage (p<0.0001), while elevated WBC count was significantly associated with skull fracture (p<0.0001) and diffuse axonal injury (p=0.0096). Additionally, elevated ED glucose and WBC count were predictive of maxillofacial fracture (p=0.0003 and p=0.0123, respectively).

In adult patients presenting acutely with mild TBI, elevated ED glucose and/or WBC count predict deleterious head CT abnormalities. Therefore, patients found to have increased ED glucose and/or WBC count should have a head CT performed promptly to assess for abnormalities.

Traumatic brain injury (TBI) is a common emergency department (ED) presentation, of which 10% can be classified as severe TBI. Deciphering early predictors may be of use in risk stratification.

This was an IRB approved observational cohort analysis of severe TBI patients (GCS≤8). Our subgroup included adult patients (18 and above) who presented during a 20 month period of a larger cohort who presented to ED and died in hospital. The studywas conducted at a level-one trauma center in the southeastern United States that is home to both emergency medicine and neurosurgery training programs. Significance level was set at p<0.05.

Death during hospitalization occurred in 38% of all the patients with severe TBI (n= 280). The mean age of these patients was 49 with 72% male. In a univariate regression model decreasing level of ED blood Glucose (p= <0.0001), lactate (p= .0045), INR (p= <0.0001), and aPTT (p= <0.0001), were significantly associated with in-hospital death. While increasing level of RBC (p=0.0002) and platelets count (p=0.0031) were significant predictors of in-hospital death. Also, younger patients had higher incidence of in-hospital death after severe TBI (p= 0.0001). Multivariate regression models showed significant association between lower ED glucose (p=0.0172), lower INR (p=0.0212), and lower aPTT (p=0.0015).

In-hospital death due to severe TBI has more impact on younger adults. Of several laboratory investigations, ED glucose, lactate, INR, aPTT, RBC, and platelet counts are potentially modifiable factors. Future studies on these predictors can be helpful to improve treatments to reduce the incidence of in-hospital death among severe TBI patients.

Patients with psychiatric disorders are at higher risk of injuries. We conducted a cohort study to investigate whether patients with pre-existing psychiatric disorders are at higher risk of worse symptoms after Traumatic Brain Injury (TBI) and Post Concussive Syndrome (PCS).

This is an IRB approved prospective cohort study included adult patients (18 and above) who came to the emergency department after a history of head injury. The data were stored in RedCap. Chi-square tests, and the z-test for proportions were done, using JMP Pro 10.0 for MAC to calculate significance between pre-existing mental disorders with symptoms and outcomes after TBI.

In the cohort of 811 TBI patients, 23% had an existing psychiatric disorder, including 14%depression, 12%anxiety, 2%Substance Abuse Disorder, 4% other psychiatric disorders (Bipolar disorder, PTSD, ADHD, Schizophrenia, OCD, Conversion disorder, and panic attack). Patients with a history of psychiatric disorders were significantly more likely to be female (p=0.0042), and sustain their head injury via MVC mechanism (p=0.0158). They were significantly more likely to experience an alteration in consciousness (p<0.0001), seizure (p=0.0013) and headache (p=0.0017). They also had a significantly higher history of prior head injury in both themselves (p=0.0005) and their family (p=0.0424). During ED evaluation with the Rivermead Post Concussion Survey Questionnaire (RPCSQ), they were significantly more likely to report blurred vision (p=0.0272), fatigue (p=0.0097), sadness/depression (p=0.0069), nervousness (p=0.0430), feeling of being slowed down (p=0.0129), and difficulty remembering (p=0.0016).

The prevalence of psychiatric disorders in the population with mild TBI is not insignificant. A history of psychiatric disorders results in a significantly higher score on the RPCSQ, which in turn, leads to poorer outcomes, including a higher frequency of post concussive syndrome.

Severe traumatic brain injury (TBI) is associated with high mortality rates and neurological disability. Management of cerebral edema and elevated intracranial pressure is a critical component when caring for brain-injured patients. Hypertonic saline (HTS) solutions have been used as a hyperosmolar therapy. In addition to osmotic, hemodynamic, and immune modulating effects, HTS plays an important role in countering hyponatremia. The aim of this study is to evaluate current practices of HTS utilization in severe TBI patients to later develop a nursing-driven titration protocol.

A retrospective chart review was conducted for patients identified between June 2011 and May 2013, who received HTS for severe TBI. Patients with severe head injury were included in the study. The primary objective was to evaluate time to achieving serum sodium goal. The secondary objectives were to assess the percentage of time patients remained in sodium goal, length of hospital stay and in-hospital mortality. For continuous variables measures of central tendency (e.g. mean, median) and standard deviation were provided. Proportions were used for categorical variables. Factors predicting time to achieving sodium goal were identified using linear regression.

Average age was 42 years (SD = 20.1), with 66% males. Overall average baseline sodium level was 140.2 (SD = 4.7). Median GCS at admission was 3 (range: 3-9). Median time to achieving sodium goal was one day (range: 0-5 days) and 9 patients remained within sodium goal (28%). Linear regression showed that shorter time to achieving sodium goal was significantly associated with increased baseline sodium level (PE = -0.13, p = 0.004) and marginally related to increased GCS on admission (PE = -0.20, p = 0.098).

The results of this study will provide data on optimal dosing regimens to achieve desired sodium goals and to develop a standardized protocol for HTS administration in severe TBI patients at our institution.

Brain activity during the early stage of traumatic coma remains unexplored. We report pilot functional MRI (fMRI) and EEG Event-Related Change (fEEG) results from the Traumatic Coma RESPONSE study, which aims to detect latent brain activity and predict recovery of consciousness.

Six patients admitted to the NeuroICU and Surgical ICU with traumatic coma were enrolled (4M, 2F; ages 19-34; GCS score 3-8), as well as four controls (2M, 2F; ages 21-33). Cases and controls underwent fMRI/fEEG with language, music, and motor imagery paradigms. Brain network activation was graded as absent, partial, or complete for each fMRI/fEEG paradigm, with expected network activation patterns defined according to prior studies. Level of consciousness was assessed at the time of fMRI/fEEG and hospital discharge using the Coma Recovery Scale-Revised (CRS-R).

fMRI/fEEG was performed on post-injury day 1-17 and was feasible in all cases despite medical comorbidities. CRS-R evaluation at the time of fMRI/fEEG revealed coma (n=1), vegetative state (n=2), minimally conscious state (n=1), posttraumatic confusional state (n=1), and full orientation (n=1). All controls exhibited complete activation of expected brain networks during each fMRI paradigm. Cases exhibited a range of fMRI brain network activity (absent, partial or complete) corresponding to levels of consciousness at the time of fMRI/fEEG and discharge. One case demonstrated partial activation of brain networks during language, music, and motor imagery stimuli, despite the diagnosis of vegetative state on CRS-R examination. While resting alpha power was lower among cases than controls, fEEG paradigms enabled analysis of activation upon language, music, and motor imagery stimuli. Language stimuli produced a trend toward increased EEG power in fast frequencies compared with alternating periods of rest.

fMRI/fEEG is feasible during early traumatic coma and may reveal clinically undetectable residual brain activity. Continued enrollment will aim to determine whether fMRI/fEEG activation patterns predict long-term functional outcomes.

Ascending reticular activating system (ARAS) lesions are implicated in the pathogenesis of traumatic coma, but methods for assessing ARAS structural and functional integrity are lacking. We performed a multimodality assessment of ARAS arousal networks during early traumatic coma utilizing advanced MRI and EEG techniques.

Six traumatic coma patients underwent MRI and EEG on post-injury day 1-17 (4M, 2F; ages 19-34; GCS score 3-8), as well as four controls (2M, 2F; ages 21-33). ARAS structural connectivity was assessed with high angular resolution diffusion imaging (HARDI) tractography by identifying fiber tracts connecting the midbrain reticular formation (MRF) to the thalamus. ARAS disconnections were assessed by quantifying ARAS microbleeds on susceptibility-weighted imaging (SWI). EEG data were assessed for reactivity to auditory/tactile stimulation and for functional connectivity using 1-second correlation networks. Clinical arousal was defined by the arousal score on the Coma Recovery Scale-Revised (CRS-R). Spearman correlations between HARDI, SWI, and EEG data and CRS-R arousal score at the time of MRI/EEG and hospital discharge were assessed.

The number of ARAS fiber tracts connecting MRF to thalamus correlated with CRS-R arousal score at MRI/EEG (R=0.9, p<0.01) but not at hospital discharge (R=0.6, p=0.18). ARAS microbleeds correlated inversely with CRS-R arousal score at MRI/EEG (R=-0.9, p<0.01) but not at discharge (R=-0.4, p=0.38) . EEG reactivity was present in all cases, including one in coma and two in vegetative state; a varied stimulation battery was necessary to elicit reactivity. EEG network density histograms suggested lower-density networks in cases compared to controls.

HARDI tractography measurements of MRF-to-thalamus connectivity and SWI measurements of ARAS microbleeds are potential "MRI biomarkers" of ARAS structural integrity in early traumatic coma. EEG may detect ARAS arousal activity even in patients without evidence of arousal on bedside examination. Future exploration will assess traumatic coma recovery in relation to EEG network density and connectivity pattern.

Use of brain tissue oxygenation (pBtO2) is a controversial variable in the monitoring and treatment of severe traumatic brain injury (TBI). We evaluated the strength of correlations between pBtO2 values and intracranial pressures (ICP) obtained from intraventricular (EVD) and intraparenchymal (IPM) monitors to assess if pBoO2 is a surrogate for ICP or if it provides unique information for the treatment of severe TBI.

Retrospective observational study of time-indexed ICP and pBtO2 values recorded for patients between October 2010 and October 2012 diagnosed with severe TBI. Indications for monitoring were by Brain Trauma Foundation Guidelines.

Only patients with simultaneous EVD, IPM and pBtO2 monitors with time-indexed charted data were eligible for inclusion.

Subjects (N=27) were primarily male (81%) with a median age of 30 years (range 13-62). There were 1,617 time-indexed data points (pBtO2, EVD-ICP, IPM-ICP). Patients were monitored for a median of 71 hours (12-216 hours). The pBtO2 values had a moderate negative correlation with both ICP values obtained from IPM (-0.29, p<0.001) and EVD (-0.27, p<0.001). There were 360 time-indexed data points with recorded ICP > 20 mmHg. The pBtO2/ICP negative correlations were stronger when ICP was > 20 mmHg for the EVD (-0.37, p<0.001) but not the IPM (-0.27, p<0.001). In 127 instances (35.2%) ICP elevations were preceded by a decreased pBtO2 one hour earlier (Chi-square 4.173, p<0.04).When pBtO2 was <15 mm Hg, the ICP was <20 mmHg in 57.1% of EVD recorded values and 32.5% of IPM recorded ICP values.

Brain tissue oxygenation monitoring provides unique information. It moderately correlates and precedes elevations in ICP and is able to detect changes even when ICP is <20 mmHg. Further ongoing clinical trials will be needed to evaluate the clinical significance of these correlations and establish if pBtO2 monitoring can improve outcomes.

Background: TBI is becoming a more and more common emergency department (ED) presentation, and has many reported sequelae. Towards this end, many types of testing in the acute setting are being investigated. One of these is convergence insufficiency (CSI) testing. CSI is a condition in which a patient finds it difficult to maintain alignment of the eyes on a near object. Inability to sustain convergence may cause a person to look with just one eye at a time, or to see double. These are common problems reported by patients with TBI, but such oculomotor testing is rarely done in the ED. Objective: We wanted to study the feasibility of doing CSI testing in the ED, and investigate whether the patient is more likely to experience post concussive symptoms, and require repeated visits to the hospital.

Methods: Written informed consent was obtained from patients age>18 who experienced a head injury from any mechanism. Patients underwent CSI using a standardized instrument of 14 questions, with responses based on the Likert scale. These data were correlated to outcomes of hospital admission, occurrence of post-concussive symptoms, and 30day hospital re-admission

A total of 104 patients were prospectively enrolled, of which 56 were men. The median age was 31, range 18-95. The median CSI score was 13, IQR 6-21, range 0-53. Men had a higher median CSI score (14 vs. 10 for women). The higher the CSI score, the more likely they were to be admitted to the hospital (p=0.0435), develop symptoms of post concussive syndrome (p=0.0721) and be readmitted within 30d (p=0.0098).

CSI can be a solid adjunct in the evaluation of acute head injury in the ED. It is easy and fast to administer, and useful to risk stratify patients in terms of who is at highest risk of developing complications.

Study objective is to determine the association between age and hospital outcomes following TBI. Health outcomes following TBI are emerging fields of research.

This is a retrospective chart review of patients presenting to the ED following a TBI. Outcome variables were: return to the ED within 72 hours of index visit, admission to the hospital, admission to the ICU, length of stay in the ICU, readmission within 30 days following discharge, in-hospital death, and death from any cause within 3 months following the index visit. Predictors were age in years and TBI severity.

The 

The data suggests that young age is associated with poorer long-term outcomes following TBI. Younger age groups tend to have more violent mechanisms of injury such as traffic accidents and assaults.

Study objective is to determine if glucose and WBC associations with abnormal head CT differ based on time of presentation to the ED following injury. Previous research has examined whether lab values are predictive of abnormal CT findings, but little research has looked at whether there is a difference based on delays in presenting to the ED.

This is a retrospective observational chart review of patients who presented to the ED following a TBI. The outcome variable was abnormal head CT. Predictors were glucose and WBC. Data was filtered to examine the entire cohort, patients who presented to the ED within 12 hours, and patients who presented to the ED after 12 hours. Analyses were conducted controlling for age of patients and TBI severity.

The cohort included 2629 patients. 2301 patients presented to the ED within 12 hours of their injury, and 328 presented longer than 12 hours after their injury to present to the ED. Low glucose(<130mg/dl) (p=0.0002) and low WBC (<14000 cells/mcL) (p<0.0001) were associated with no abnormal head CT when looking at the entire cohort. Low glucose (p<0.0001) and low WBC (p<0.0001) were also associated with no abnormal head CT in patients who presented to the ED within 12 hours of their injury. Neither glucose nor WBC was associated with no abnormal head CT in patients who presented to the ED longer than 12 hours following their injury.

Low glucose and low WBC are associated with no abnormal head CT findings in TBI patients who present to the ED within 12 hours of injury.

Study objective is to determine any association between Aspirin or Warfarin use and component GCS scores. Aspirin and Warfarin are known to increase the risk of excessive bleeding. Little research has focused on how this impacts component GCS scores.

This is an observational cohort study of patients who presented to the ED following a TBI. Outcome variables were GCS score and each of its components: eye, verbal, and motor scores. Predictors were Warfarin and Aspirin use. Analyses were controlled for age.

The cohort included 2633 patients. 2243 had mild TBI, 82 had moderate TBI, and 308 had severe TBI based on GCS scores. 436 patients were on Warfarin when they presented to the ED of which 331 were mild TBI patients. 169 patients were on Aspirin when they presented of which 153 were mild TBI patients. Aspirin use was associated with higher overall GCS (p=0.0088), higher eye score (p=0.0068), higher verbal score (p=0.0054), and higher motor score (p=0.0088) when looking at the entire cohort. Warfarin use was associated with lower motor score (p=0.0214) in mild TBI patients. Neither Aspirin nor Warfarin use was associated with eye or verbal scores in mild TBI patients.

Aspirin use is associated with higher GCS scores when examining the TBI population as a whole. Warfarin use is associated with lower motor scores in mild TBI patients. Neither appears to be associated with independent eye or verbal scores in the mild TBI population.

This study will compare TBI severity and outcomes between men and women taking anticoagulants.

This observational cohort study of adult patients who came to the Emergency Department with a TBI. Outcome variables included ED GCS scores, neurosurgical interventions, hospital admissions, 72-hour ED returns, in-hospital deaths, death within 3 months, CT scan abnormalites. Predictors included gender and anticoagulant use.

The cohort included 2639 patients, of which there were 140 females and 148 males taking anticoagulants. Overall, patients taking anticoagulants had worse outcome than those not on anticoagulants. Within the anticoagulant group, a Fit Least Square Regression Model revealed that females had a higher GCS score (mean of 13.94 vs. 13.64) following a TBI than do males (p<0.0001). Females also had fewer 72-hour ED returns (p=0.0558), neurosurgical interventions (p=0.0001), hospital admissions (p<0.0001), deaths within 3 months of injury (p=0.0106), and in-hospital deaths (p=0.0607). Females on anticoagulants are also less likely to have CT scan abnormalities (p<0.0001), and when there is an abnormality on the CT scan, they are less likely to have a skull fracture (p=0.004). The female cohort had less bleeds on CT, though not statistically significant (p=0.0808). Mechanism of injury was analyzed as a possibility for the discrepancy, and more males on anticoagulants were in motor vehicle collisions (MVCs) than females (24% vs. 16%, p=0.2648).

We observed that female patients taking anticoagulants experience less severe TBIs and have better outcomes than their male counterparts. Analysis revealed that a larger amount of TBIs occurred through MVCs for males than females, which may account for some of the discrepancy. There didn't appear to be any other sizeable differences in mechanism of injury between the genders, indicating that further studies are needed to understand why this difference exists.

Previous studies from our lab have shown that acute laboratory findings can be associated with certain short-term outcomes after TBI. This study examines ED glucose and WBC to see if they have predictive power for TBI outcomes. This study will investigate whether there is a connection between Emergency Department (ED) Glucose and WBC count and several key post-TBI outcomes.

This is an observational cohort study of adult patients (≥18) who came to the ED department with a TBI. The individual outcome variables were hospital admission from the ER, hospital length of stay, in hospital death, and 3-month death. The predictors were blood glucose levels (>140 was considered high) and WBC count (>14 was considered high).

Of the 2639 TBI patients in the cohort, 1160 were readmitted, 147 died in the hospital, and 179 died from any cause within 3 months. A regression model revealed that lower blood glucose and white blood cell counts are associated with fewer hospital admissions from the ER, a shorter hospital length of stay, fewer in hospital deaths, and fewer 3-month deaths, with p<0.0001 for all of the above findings. Further analysis was performed to see whether these relationships held for patients with severe TBIs (GCS<9), and the same relationship was found between glucose and in hospital death (p<0.0001), glucose and death in three months (p<0.0001), and WBC count and hospital length of stay (p=0.0191).

It appears that glucose and WBC count are associated with various TBI related outcomes, though WBC may not be as strong of a predictor of glucose for severe TBIs. Several other of our lab's studies have also suggested that glucose and WBC count can serve as predictors for TBI outcomes and findings, indicating that further studies are needed to ascertain why that is the case.

Past studies have raised questions about the necessity of CT scan use following mild TBI. Whether or not radiologic tests are more useful for patients on anticoagulants and antiplatelets following TBI is yet to be seen. Though guidelines like the Canadian CT Head Rules have been created for mild TBI CT scan use, many, including the Canadian Rules, have excluded anticoagulants from their criteria. This study will attempt to determine if anticoagulant use in the general and elderly population results in Traumatic Brain Injury patients having more radiologic abnormalities following injury.

This is a cohort observational study of adult patients (≥18) who came to the ED department with a mild TBI (GCS≥13).

The individual outcome variables were CT scan orders, CT scan abnormalities, and CT bleeds post-TBI. The predictors were anticoagulant use and age (age≥55).

Of the 2639 TBI patients in the cohort, 2243 patients had mild TBI (GCS≥13). Of those patients, 254 were using anticoagulants. Results showed that anticoagulant use was associated with abnormal CT findings (p<0.0001) and CT bleeds (p<0.0001). When patients greater than 55 years of age with mild TBI were considered, 214 were taking anticoagulants. We observed similar results with this population, showing that anticoagulant use is associated with abnormal CT (p=0.0077), and CT bleeds (p<0.0001). Additionally, CT scans were ordered more frequently for anticoagulant patients (p=0.0268).

We observed that patients on anticoagulants following mild TBI receive more CT scans than patients not on anticoagulants. This strategy appears to be validated by the results of this study, which shows that anticoagulant use is associated with an increase in radiologic abnormalities, for the general and geriatric population suffering mild TBIs. Thus, these findings suggest that there is an added need to order CT scans for mild TBI patients using anticoagulants.

The pediatric population is at an increased risk for rare but significant injuries that can occur from radiation exposure due to CT imaging. With the trend of increasing head CT use in the evaluation of pediatric traumatic brain injuries, finding identifying factors that can help prevent excessive CT usage in this vulnerable population is important. Our objective is to characterize peri-traumatic amnesia and other predictors of normal neuroimaging results in the context of pediatric traumatic brain injury.

This was an IRB approved observational cohort study in which children (age≤17) were enrolled if they presented to the ED with a head injury. Data was collected into RedCap, and analyzed with JMP 10 for Windows.

Of the cohort, 70% underwent a head CT imaging study with 31% having abnormal findings. Common abnormal findings included calvarial fracture (15%), Parenchymal/hemorrhagic contusions (7%), and extracalvarial swelling (16%). In our regression analysis, we found that several factors were significantly correlated with normal CT findings in patients that underwent imaging studies. Patients who denied post-traumatic amnesia (p=<.0001), alteration of consciousness (p=<.0001); and who did not experience loss of consciousness of less than thirty minutes (p=<.0001), were unlikely to have abnormal head CT.

While positive neurological findings are not uncommon in pediatric head injury, a large percent of pediatric patients who undergo scans reveal no abnormal findings. While some of these factors have already been included in predictive tools such as the PECARN CT rule, lack of peri-traumatic amnesia has not been included in these rules, and may be an additional helpful adjunct.

Many children use recreational vehicles, such as bicycles, daily but required use of helmets in this vulnerable pediatric population is a controversial topic and laws are not universally in place. Studies emphasizing the importance of proper helmet use during the operation of recreational vehicles can be of influence in the policy discussion around this topic.

This was an IRB approved observational cohort study in which children ages 5-17 were enrolled if they presented to the ED with a head injury. Data was collected into RedCap, and analyzed with JMP 10 for Windows. Recreational vehicle was defined as any of the following: bicycle, motorcycle, ATV, watercraft or other (eg. horse).

In our cohort, 125 patients reported the use of a recreational vehicle prior to injury. ATVs were the most used (32.8%), followed by bicycles (24.8%), motorcycles (23%), other (16%) and watercraft (2.4%). Helmet use was assessed in 94.4% of this population. Only 12% of bicycle and 17% of ATV associated injuries reported helmet use. Seventy-five percent of motorcycle accidents reported helmet use. Eighty-seven percent of our cohort had a mild TBI (GCS=13-15), 2.4% had moderate TBI (GCS=9-12), and 10.4% had severe TBI (GCS<9). Of those with severe TBI, none reported helmet use and 61% were due to ATV accidents. Regression analysis in our population showed helmet use to be protective of hospital admission (p=.027), abnormal CT findings (p=.0002), and calvarial fracture (p=.0012).

These data support the importance of helmet use in a pediatric population as a protective device. The majority of states do not have laws requiring helmet use on recreational vehicles, and our data help to support policies requiring use of this protective measure.

Every year upwards of 180,000 children are involved in traffic accidents (TA). This mechanism of injury places them at risk for both acceleration/deceleration injuries along with mechanical trauma. Our goal is to characterize traumatic brain injury in pediatric patients presenting with traffic accidents as the mechanism of injury.

This was an IRB approved observational cohort study in which children (age=<17) were enrolled if they presented to the ED with a head injury as identified by ICD-9 codes. Data was collected into RedCap, and analyzed with JMP 10 for Windows.

In our cohort (n=1173), 185 (15%) presented following a traffic accident. 19% were reported as drivers, 57% were passengers, only 3%( n=5) were in open air seating and 21% were unknown or other. Grouped by GCS scores, 30 had severe TBI (<9), 9 had moderate TBI (9-12), and 146 had mild (13-15). 51% of severe TBIs in our cohort were involved in TA. 63 patients had abnormal CT findings, while 11 required surgical interventions, 33% of all surgical interventions in our cohort. Seatbelt usage was collected for the cohort, with 38% of patients were unrestrained, 43% were restrained, and seatbelt use was not identified in 19% Use of restraint was significantly correlated with normal CT findings (p=.0066), no loss of consciousness (p=.0002), no alteration of consciousness (p=.0002), peri-traumatic amnesia (p=.0153), higher GCS score (p=.0013) and discharge from the emergency department (.0074).

WhileTA is not the most common cause of TBI in the pediatric population, they accounted for high percentages of severe TBI and surgical interventions in our cohort. Our findings emphasize the importance of properly restraining a child traveling in a motor vehicle, as use of a seatbelt is significantly associated with less morbidity in patients with TBI.

Imbalances in baseline factors frequently occur in early clinical trials in stroke and traumatic brain injury (TBI) resulting in uncertainty in how representative they would be of likely outcomes in a larger population. Traditional multi-variate correction methods cannot accurately account for non-linear relationships and frequently fail to accurately predict the result of Phase 3. We sought to test whether a method originally developed to use in stroke to predict success/failure of early Phase trials could be adapted to TBI. pPREDICTS © (Stroke, 2009) generates a surrogate outcome model based on the placebo arms of randomized clinical trials using baseline factors that influence outcome. The novel feature is the generation of multi-dimensional statistical surfaces above and below the outcome function to test how a trial would compare to this more representative population. We tested the relationship between baseline Glasgow Coma Scale (GCS), age, mortality and functional outcome.

Twenty-eight placebo arms or consecutive case series were found with adequate data on baseline GCS, age and outcomes (mortality, unfavorable, and favorable) at > one month. pPREDICTS-TBI Matlab program was developed to perform a non-linear fit and provide ±95% prediction interval surfaces (p<.05; p>.05).

The selected studies represented 7342 subjects. The predictive power of these models was excellent (R 2 : Mortality=0.67, Unfavorable outcome=0.81, Favorable=0.87; all p<0.0001).To test whether this model confirmed that outcomes from TBI improved over time, outcomes from 1985 showed mortality above the p=.05 statistical interval and mortality from a 2013 trial were below the p=.05 surface.

We show that multi-dimension outcome models along with prediction intervals can be developed for TBI. Early trial/case series outcomes can be compared against the models, and placebo arms can be tested for how representative it is of a larger population. This method may aid in selection of trials that should move forward to later Phase.

The coagulopathy associated with traumatic brain injury is comprised of both a hypo-and hypercoagulable phenotype. Controversy exists regarding the sequence of bleeding and thrombotic manifestations, as well as the duration of coagulopathy. We sought to determine whether thromboelastography could identify a late hypercoagulable state after TBI, defined by elevations in TEG parameters: maximal amplitude (MA), alpha angle (αA), G value (G) and total thrombin generation (TTG).

Patients were prospectively enrolled in a cohort study. Patients with traumatic brain injury and GCS <12 (or higher admission GCS but >2-point fall) admitted between 1/2012 and 12/2012 were eligible for enrollment. TEG profiles were obtained between 24-48 hours (T1) after initial injury and 72-96 hours (T2) after initial injury. 

The data suggest that there is a delayed hypercoagulable state observed after TBI. The hypercoagulable state may reflect excessive platelet activity. However, the exact mechanism underlying the hypercoagulable state remains elusive.

Background:The Glasgow Coma Scale (GCS) was developed as an objective tool to measure level of consciousness in traumatic brain injury (TBI).It is the cornerstone of clinical assessment in severe TBI and robust prognostic value in TBI models (CRASH and IMPACT). Studies have shown a high degree of inter-and intra-rater reliability of the scale from observers having an array of experience, however according to Waterhouse 2005, the GCS is "misunderstood and misused". PURPOSE To describe the results of an educational review effort provided to nurses working in a Trauma and Neurosurgery (TN) Intensive Care Unit, TN In-Patient Unit and Critical Care Rapid Response Team (CCRT) in a level-one trauma centre in a teaching hospital in Canada.

A thirty-minute PowerPoint session was presented to a group of nurses that described the components of the neurological observational record (GCS, pupil, limb movement, vital signs). The interactive session was based on the components and clinical case scenario's nurses encounter in daily practice; multiple-choice questions along with the IClickers audience response system guided discussion.

Seven sessions were provided which ninety-one nurses (fifty-eight TNICU, twenty ward and thirteen CCRT) participated. A discrepancy between current literature and practice was identified. A mean average of 64% of nurses could accurately interpret and document neurological findings (90% could identify central pain responses whereas 21% could identify an acceptable form of a central pain stimulus). Because IClickers collected answers anonymously, we could not identify variability between the difference groups of nurses.

Findings from this pilot study identify the need for neurological observational training as a result of the misinterpretations and misunderstandings to accurately perform a consistent neurological assessment. Teaching sessions can be provided during unit orientation for new hires and on a continual basis for existing staff.

To determine which post-concussion symptoms most commonly appear 3-15 days after a motor vehicle accident.

The observational prospective study at a level-one trauma center included 811 patients who arrived within 24-hours of a head injury. This study examines the subset involved in automobile or motorcycle accidents. Patients received follow-up through phone calls. Patients were assessed for various symptoms, including headache, nausea/vomiting, dizziness, tinnitus, sensitivity to light/noise, vision changes, numbness/tingling, drowsiness, fatigue, emotional changes, sleep disturbances, and difficulty concentrating or remembering. Data were entered into RedCap and analysis conducted with JMP 9.0 for Mac.

For a cohort of 405 patients involved in motor vehicle accidents,58.91% report at least one post-concussion symptom at the time of phone follow-up, as compared to 44.83% for patients who report at least one symptom as a result of alternative mechanisms (p-value = 0.0046, OR = 1.76, 95% CI 1.19-2.61). Contingency analysis was used to assess the probability of each symptom, and statistically significant findings include headaches (p-value = 0.00249), tinnitus (p-value = 0.0402), vision changes (p-value = 0.0261), numbness/tingling (p-value = 0.0033), and sleeping more than usual (pvalue = 0.0290). Difficulty in concentration was not statistically significant, but more common in MVA patients. Multivariate regression showed that when controlling for age, MVA patients are significantly more likely to exhibit post-concussion symptoms.

Patients who are involved in motor vehicle accidents are more likely to have experienced post-concussion symptoms after a traumatic brain injury such as headaches, tinnitus, vision changes, numbness/tingling, sleeping more than usual, and difficulty concentrating.

Financial Support: None

The major objective during the acute care of traumatic brain injury (TBI) is the avoidance of secondary brain injury. This is achieved, in part, by ensuring adequate cerebral blood flow and oxygen delivery and so it seems physiologically rational to transfuse packed red blood cells (PRBCs) when TBI patients are anemic. However, there are several lines of evidence that suggest that transfusion of PRBCs in critically ill patients may be harmful. Thus, there is a diverse opinion on the benefits and thresholds for transfusion in TBI and an urgent need to understand the neurophysiologic effects. This study aims to provide objective, non-invasively achieved, physiologically relevant data in order to provide some rational basis for decision-making for transfusion in TBI patients.

We have prospectively enrolled five TBI patients admitted to the Neuro-ICU who, at the discretion of the treating physician, reached a threshold for transfusion of PRBCs. We evaluated the applicability of four-wavelength near-infrared spectroscopy (NIRS) to monitor the cerebral tissue oxygenation (StO 2 ) in TBI patients during a transfusion of PRBCs. We present the preliminary results testing the hypothesis that the transfusion of PRBCs will improve oxygen delivery and will be associated with improved StO 2 values.

Five patients were measured over a 24-hour period. StO 2 values strongly correlated with values obtained from the contralateral side (p <0.01). StO 2 values increased after a transfusion but this did not reach statistically significance in all patients.

These preliminary results outline a novel procedure to evaluate the cerebral tissue oxygenation in traumatic brain injured patients during a blood transfusion. The results illustrate to what degree cerebral tissue oxygenation is dependent on the degree of anemia and its correction. The results did not reach statistical significance for all patients, this may be due to a lack of cerebral autoregulation, increasing global brain dysfunction or storage lesions in PRBCs.

Hypertonic (3% NaCl) solutions are commonly used to manage acute intracranial hypertension and cerebral edema. Even when central lines are not available, peripheral infusions of 3% NaCl (1027 mOsm/L) are often avoided because of concerns for infiltration and phlebitis. Medical centers often limit 3% NaCl infusion rates (≤ 30mL/hr) when administered peripherally. To assess the safety and feasibility of peripheral infusions of 3% NaCl, we studied the incidence and risk factors of infiltration at rates up to 100mL/hr.

We prospectively evaluated adult patients that received 3% NaCl infusions for >24 hours at rates up to 100mL/hr. Our primary endpoint was catheter infiltration (INFILT), defined as inadvertent administration of a nonvesicant solution or medication into the tissue surrounding the IV catheter. Subgroup analysis was conducted to examine if infiltration was affected by infusion duration, age of the catheter (from insertion until the start/end of infusion), catheter size, and maximum catheter infusion rates ≤ 30 (LOW RATES) vs > 30mL/hr (HIGH RATES).

Infiltration occurred in 26 of 84 catheters (31%), but none required surgical intervention. The median infusion duration in the INFILT group was shorter than the group without infiltration (INTACT), (29 vs 39 hours, p=0.29). The median age of the catheter before initiating 3% NaCl was less in the INFILT group than the INTACT group (2.4 vs 7.2 hours, p=0.28). Infiltrations were not affected by catheter size. There was no difference in the incidence of infiltration with LOW RATES (16 of 54 catheters, 30%) vs. HIGH RATES (10 of 30 catheters, 33%), p=0.81.

Infiltration rates were high (31%) with peripheral infusions of 3% NaCl, but were not affected by catheter age, catheter size, or maximum infusion rate. To minimize infiltration risk with prolonged 3% NaCl peripheral infusions, consider switching to a central line within 24 hours.

The use of thromboelastegram (TEG) to determine presence of coagulopathy in a variety of surgical and non-surgical patients is well established and used commonly to guide transfusion of blood products. TEG guided chemoprophylactic therapy to decrease the occurrence of thromboembolic (TE) events following traumatic brain injury (TBI) is less well studied. The purpose of this literature review is to provide a summary of the evidence supporting the utility of TEG-guided chemoprophylactic therapy in the TBI.

Comprehensive literature review using PubMed and the search terms TEG/Thromboelasegram, Traumatic Brain Injury, and Thrombus for the years 2003 -2013. The search resulted in 1415 articles, of which there were 18 systematic reviews in human subjects. 11 articles discussed TEG and trauma.

Eleven articles were selected for this review summarizing results from 9,481 patients. Six articles discussed TEG profiles to assess platelet dysfunction, coagulopathy, and hyperfibrinolysis. Three reported decreased bleeding associated with utilizing TEG to assess for coagulopathy following TBI. Two of the studies reported a decreased use of blood products after TBI when TEG profiles were used to guide therapy. In a single trial, abnormal TEG was correlated with mortality. One systematic review of TEG used to predict TE events postoperatively concluded that the predictive accuracy for TEG is highly variable due to variations in hypercoagulability definitions and TEG methodology but suggested that maximum amplitude (MA) seems to be the best parameter to identify TE events.

Coagulopathy and platelet dysfunction are increasingly recognized as having a direct association in TBI and to overall clinical outcomes. TEG has distinct advantages over traditional plasma based coagulation tests. There is adequate evidence to support the assumptions required to design a prospective randomized clinical trial of TEG for the prevention of thromboembolic events after TBI.

Minor traumatic brain injury (mTBI or concussion) has seen changes in resources devoted to education, and awareness as well as structured limitations on athletic concerns. Few studies to date have attempted to determine whether, increased occurrence is related to change in injury patterns or improvements in physician awareness and diagnosis. Objectives: To determine if mTBI rates are increasing faster than all trauma and whether detection is related to better diagnosis or increased occurrence.

The Emergency Department and Trauma Center records were analyzed at ED and Trauma Centers in 2 metropolitan areas for the past decade 2000-2010. Trauma registries and the AZYXXI database (Microsoft; Redmond,WA) were analyzed for trauma admits,ED visits and mTBI rates and treatment interventions including use of radiographic study and dispositions. mTBI defined as arrival GCS 15, possible LOC and no other injury. IRB approval and data analysis was obtained and performed, respectively.

A 10 year study found rapid rise in past 5 year with number of concussions which increased by 140% compared to ED pt. census and Trauma pts volume increased only by 23.9%; p<0.02. (Figure 1 ) There were also increases in use of CT for concussion by 25.8% with less than 1.2% of mTBI pts. having a positive finding on Head CT and none requiring neurosurgical intervention. Both number of concussions and admitted concussions experienced the same rate of rise as number of concussions and equal AUC (area under curve).

There has been an effective impact on mTBI presentation and admission to our trauma centers in the past five years. CT increased in use with no improved treatment intervention. Future studies will need to determine utility of admit compared to outpatient observation and neuropsychiatric intervention .

Most hospitals do not have 24/7 neurosurgical capabilities and even more severe shortages of neurosurgical services in the United States are expected in the future. For rapidly expanding intra-cranial hemorrhage and deteriorating clinical status, the patient outcome is universally poor if transfer to a tertiary care institution is required for hematoma evacuation. Although opportunities to perform emergent trephination in the ED are rare, the knowledge and ability to perform a lifesaving "burr" hole is usually non-existent.

The SimCenter of a 950 bed urban, Level 1 Trauma, teaching hospital, was used for the training scenario: PGY 1-3 EM residents received patient scenarios having signs of increased intracranial pressure due to an expanding hemorrhage requiring emergent cranial trephination. -A novel skull hematoma model was developed.-Demonstration of the correct procedural steps needed for cranial trephination and extra-ventricular drain (EVD) -Pre-and post-knowledge-based surveys were completed by each resident to assess understanding of the diagnosis and management of an expanding intracerebralhemorrhage, incidence of performing trephination in the ED, and confidence level in skill performance.-Time to successful hematoma drainage was assessed both pre-and post-instruction.

-20/20 (100%) EM residents completed both sim-training and successful skull trephination and EVD placement.-Mean time for trephination and EVD placement: 5.8 min (IQR: 3.9-7.4 min).-1 of 20 residents (8%) had performed cranial trephination and 0/20 (0%) performed this procedure during their training.-Resident comfort level in performing ED cranial trephination increased from 0.3POST8(95%CI:0.23-0.50) for PRE versus 6.82(95%:5.4-7.2) POST training.-Additional sim-training increased likelihood to perform cranial trephination: 6.82 (95%CI: 5.5-8.1) and residents favored the additional simtraining on presentation, diagnosis and emergent management of an expanding intracerebralhemorrhage measured with 10 cm VAS:6.36 (95%CI: 5.7-7.0).

This reproducible simmodel provides a novel approach to teaching residents and could facilitate maintenance of skills as attending required for performance of cranial trephination.

Air travel may be associated with unmeasured neurophysiological changes in an injured brain that may impact postconcussion recovery. No study has compared impact of commercial airtravel on concussion injuries despite rather obvious decreased oxygen tension and increased dehydration impact on acute mTBI.

Prospective cohort study of all active-roster NFL and National Hockey League players during the 2010-2011 seasons.

Internet website review of League sties for injury identification of concussive injury and when player returned to play solely for mTBI. Team schedules and flight times were also confirmed to include only players who flew immediately following game (within 4-6 hr).

During the 2010-2011 NFL and NHL seasons, 122(7.2%) and 101(13.0%) players experienced a concussions ( percent of total players), in respective leagues . Of these, 68 NFL ( 57%) and 39 NHL (39%) flew within 6 hours of the incident injury. The mean number of games missed for NFL and NHL players who traveled by air immediately after concussion was increased by 29% and 24% (respectively) than those who did not travel by air NFL: 3.8 (SD 2.2) VS. 2.6 games (SD 1.8) and NHL :16.2 games (SD 22.0) vs.12.4 (SD 18.6); p <0.03.

This is initial report of an increased rate of recovery in terms of more games missed, for professional athletes flying commercial airlines post mTBI compared to those that do not subject there recently injured brains to pressurized airflight. The obvious changes of decreased oxygen tension with altitude equivalent of 7,500 feet, decreased humidity with increased dehydration and duress of travel accompanying pressurized airline cabins all likely increase the concussion penumbra in acute mTBI. Early air travel post concussion should be further evaluated and likely postponed 48-72 hr. until initial symptoms subside.

Moderate-severe traumatic brain injury (msTBI) remains the leading cause of death after trauma. Withdrawal of care (WOC) has been identified as the leading cause of death for all neurological emergencies, and premature WOC may result in self-fulfilling prophecies. While WOC rates in msTBI range between 45%-87%, factors associated with WOC remain unknown. The objectives of our study were to determine factors independently associated with WOC in msTBI patients, and how WOC may affect mortality and surgical interventions in these patients.

Rates of WOC and predictors of in-hospital mortality, WOC, and receipt of neurosurgical interventions were determined in 232 prospectively collected msTBI patients consecutively enrolled between November 2009 and March 2013 in the ongoing Outcome Prognostication in Traumatic Brain Injury (OPTIMISM) Study at a level-1 trauma center. Multivariable logistic modeling was utilized to adjust for significant predictors in the univariate analyses.

The mean age of study patients was 53 years, with a median Glasgow Coma Scale 6 (IQR 8;3). The in-hospital casefatality rate was 36%, and 68% underwent WOC. Factors independently associated with WOC were advanced age, admission pupillary reactivity, ICU length of stay, pre-admission cardiac arrest, admission INR elevation >1.5, cardiac arrest in the ICU, herniation and ICP crisis (model C-statistic 0.93). After adding pre-existing comorbidities to the model, INR elevation on admission was no longer significant, but history of endocrine disease was. Inclusion of WOC in the model predicting in-hospital mortality negated all other variables. Craniotomy was not associated with WOC, but ICP monitor placement was.

In our msTBI cohort, WOC is the most important predictor of in-hospital mortality. Additionally, we identified important predictors of WOC. Awareness of these predictors may be important to prevent clinical nihilism in these patients.

Takotsubo cardiomyopathy (TCM) is a form of stress-induced cardiomyopathy that occurs following various neurological disorders, but there is paucity of literature on the occurrence of this entity following traumatic brain injury (TBI). Our aim is to describe the clinical outcomes of TCM in TBI patients.

We identified patients with TBI and TCM in the Nationwide Inpatient Sample Database for 2009 and 2010. We compared patient demographics, hospital characteristics and outcomes between the TCM and non-TCM groups. The primary outcomes were in-hospital mortality and home discharge. Secondary outcomes were length of stay (LOS) and total hospital cost. We used Pearson's Chi-square test and Wilcoxon-Mann Whitney tests for categorical and continuous variables respectively. We also built multivariate logistic regression models to assess TCM outcomes after adjusting for potential confounders. 

TCM is associated with worse outcomes in TBI patients, with lower home discharge rate, longer length of stay and higher hospital costs. While TCM does not increase mortality in other neurological disorders, it adversely affects mortality in TBI. The low incidence of TCM with TBI is perhaps due to under diagnosis. Prospective confirmation is warranted.

The current approach to monitoring patients with any type of acute severe brain injury is based on the insertion of multiple probes into the brain parenchyma. Over the past several years we have been working on developing a single device that can monitor several physiological parameters. A prototype of such a device has been developed and in this study we studied five modalities (temperature, oxygen, glucose, lactate and electrical activity) using the smart catheter in a rat stroke model. We report very interesting correlations between these parameters, especially as they relate to the phenomenon of spreading depression (SD).

Microsensors for the five modalities were fabricated using microelectromechanical technology (diameter=0.5mm). Electrocorticography electrodes consisted of heterostructured platinum/iridium oxide to enhance compatibility with fullband recordings. To assess the smart catheter, it was inserted into the parietal cortex of anesthetized rats that were subjected to permanent middle cerebral artery occlusion (MCAO) on the same side as the recording device.

During the 3 hours after MCA occlusion, spontaneous waves of depolarization appeared in the ischemic penumbra. The first episode appeared 2.9±1.3 min after occlusion. SDs consisted of negative DC deflections, followed by positive waves (4.7±0.9 mV). After a delay of 1.8±1.1 min from the onset of the negative DC shift the smart catheter recorded increased brain lactate (24±17 μM) and temperature (0.21±0.06 ºC); and decreased glucose (34±11 μM) and oxygen (2.9±1.8 mmHg).

With a single device, we found changes in cerebral glucose, lactate, oxygen, and temperature that correlated with SDs. These results demonstrate that the smart catheter is capable of simultaneously and continuously measuring multiple brain variables, within the pathophysiologic ranges observed in brain injury. This technology has the potential to eventually be used for the study of pathophysiology and in the acute monitoring of the injured brain.

There is controversy regarding brain tissue oxygen monitors (BTOM), in particular their optimal placement and what they actually measure. Recent work{Jaeger2010} suggests that brain tissue oxygen (PbtO2) parallels suboptimal CPP in TBI as defined by vasoreactivity index (PRx); at the minimum it gives support to the concept of optimal CPP (CPPopt){Aries2012} as an actionable target. We explore the relationship of PbtO2 with automated CPPopt in the negative deltaCPP range (underperfusion).

18 TBI patients with BTOM were studied. Post-catheter CT scans were reviewed for tip location and classified by position:

(1) normal-appearing brain;

(2) normal-appearing brain but patient had DAI; (3) peri-contusion (modified from Ponce{Ponce2012}, category added for DAI). CPPopt was calculated from a moving 4-hour window-analysis of PRx and CPP. DeltaCPP (CPPact-CPPopt) was calculated for each 60s-meanCPPact value relative to previous 4-hour CPPopt.

Positive deltaCPP values were removed for analysis focusing on negative deltaCPP values. Correlation coefficient was calculated for each subject; mean coefficient was calculated for each tip location category. ANOVA was performed comparing correlations of PbtO2 and negative deltaCPP between tip locations.

4 subjects' tips were in normal-appearing brain; 6 subjects' tips were in normal-appearing brain but DAI present; 8 subjects' tips were peri-contusional. Because r values were generally <0.5, we did not perform Fisher transform. Mean r values were 0.044 for position 1, 0.190 for position 2, and 0.020 for position 3. ANOVA of mean coefficients for each tip location category yielded a non-significant p-value of 0.54.

PbtO2 had a weak to negligible correlation with negative deltaCPP as defined by automated, continuous CPPopt calculation, regardless of tip location. This seems to contradict Jaeger's finding of BTOMs' ability to reflect suboptimal CPP; however, their work was based on summarized data while our approach focuses on prospective real-time calculations, applicable in patients' management.

Introduction traditionally, neuropsychometric testing is performed weeks to months after head injury, and mostly in patients who continue to have symptoms or difficulties. In this study we sought to determine whether these tests, when administered acutely could assist in predicting short-term outcomes after acute traumatic brain injury (TBI). The objective is to determine whether neuropsychometric testing for TBI in the acute setting has any association with acute outcomes.

Patients age>18 with a head injury from any mechanism that occurred within 24 hours of presentation to the Emergency department (ED), were enrolled prospectively after written informed consent and took 2 separate neurocognitive tests, the Galveston Orientation Amnesia Test (GOAT) and the Rivermeade Post Concussion Survey Questionnaire (RPCSQ). The GOAT is a 20-question instrument that is scored from 0-100, and the RPCSQ is a 30-question instrument with a score range from 0-65. Independent variables included raw scores on each of these tests; dependent variables included hospital admission, development of post concussive syndrome (PCS), and 30-day readmission rate. Statistical analyses were performed in JMP 10.0.

The median GOAT score was 99 (IQR 98-100, range 84-100). Having a lower GOAT score was significantly associated with being hospitalized (p=.0139), and developing post concussive syndrome at 30-45 day follow-up (p=0.0183 R 2 = 12.2%).The median RPCSQ score was 12 (IQR 5-23, range 0-61). A higher RPCSQ score was significantly associated with hospital admission (p=.0113), re-admission to hospital within 30 days (p=0.0019) and evidence of PCS at day 3-15 post injury (p=.0001, R2= 22.6%).

While not commonplace, neuropsychometrictesting in the ED in the setting of acute head injury is both feasible and appears to have value in predicting hospital admission, and who will suffer from PCS. These data are especially important in terms of helping patients understand what to expect, which in turns aids in their recovery.

Background: Head injury and mild TBI are increasingly becoming more common presentations to our EDs. Anecdotally it appears that many of these patients also have underlying sleep problems, anxiety and/or depression. The current study investigates the frequency of these disorders in our TBI population.

Methods: Patients were prospectively enrolled via written informed consent, including ED participation and telephone follow-up on day 3. The Epworth sleepiness scale (ESS), a 0-24 scale, was used to assess for sleep disorders. A score of >10 ESS is considered "at risk" for a sleep disorder based on excessive daytime sleepiness behavior. Psychiatric illness was assessed with a standardized survey instrument specifically asking about the presence of anxiety, depression, psychoses, or other mental illness. Data were entered in blinded fashion into RedCap and analyzed using JMP 10.0 Pro.

Results: The median age of the cohort was 26. The median ESS was 7.5 (IQR 5-11, range 0-16). 44% were male. 12% sustained multitrauma. The median Injury severity score (ISS) was 17 (IQR 6-37). Psychiatric history was significant for 17% depression, 12% anxiety, 3% substance abuse, 4% other psychiatric illness. A history of depression was significantly associated with multitrauma (P=0.0557) and higher ISS (P=.0446). Similarly, a history of substance abuse was also associated with a higher ISS, as well as a higher 30-day hospital re-admission rate (P=0.0006). There was a trend towards statistical significance for higher risk of having persistent concussive symptoms in those with a history of substance abuse (P=-.055).

Conclusion: The prevalence of psychiatric and sleep disorders is patients presenting to the ED with head injury is not insignificant. Furthermore, these disorders are associated with a higher risk of sustaining multitrauma and higher ISS. It is thus important to solicit this history in TBI patients, both for the acute phase as well as for follow-up.

Objectives: To identify predictors of outcome in patients with Traumatic Brain Injury (TBI) due to mechanism of fall.

This was an observational cohort study of consecutive adults who sustained a head injury within the prior 24h. It conducted at aLevel-1 Trauma center over 20 months. Data were entered into RedCap, and JMP 10 for used to perform statistical analyses.

The cohort (n= 612) had the following symptomatology: Vomiting-8%; Seizure-3%; LOC-50%; LOC > 30 min-17%; AOC-29%; post traumatic amnesia-25%. 20% were on an anticoagulant (AC)/or antiplatelet (AP) agent. Outcomes: 47% had an abnormal head CT, 43% were admitted to the hospital, 26% had an ICU stay, 8% required neurosurgical intervention, 5% died in hospital, 9% were re-admitted within 30 days, and 8% were dead at 3 months. 

Patients with symptoms such as alteration of consciousness and post-traumatic amnesia after traumatic brain injury as a result of fall are more likely to be admitted to ICU with significantly longer ICU length of stay. Mild traumatic brain injuries in fall patients should not be overlooked in daily practices because of significant morbidity and mortality.

Evidence is emerging that isolated traumatic subarachnoid hemorrhage (ITSAH) may be a milder form of traumatic brain injury (TBI). If true, ITSAH may not benefit from ICU admission which would in turn decrease resource utilization. We compared the presentation and clinical course of subjects with ITSAH to all other TBI, and performed descriptive statistics on the subset of ITSAH subjects presenting with a Glasgow Coma Score (GCS) of 13-15.

Retrospective review of a prospectively maintained database created to track the radiographic findings and clinical courses of all patients with traumatic intracranial hemorrhage from 2/2010 to 11/2012.

Of 698 subjects, 102 had ITSAH and 596 had any other intracranial hemorrhage pattern. When compared to all other subjects with TBI, ITSAH had significantly lower ISS scores (p<0.0001), lower head AIS scores (p<0.0001), higher ED GCS (p<0.0001), shorter ICU lengths of stay (p=0.007), higher discharge GCS (p=0.005), lower mortality (p=0.003), and underwent significantly fewer head CT scans (p<0.0001). In considering those ITSAH subjects who presented with a GCS of 13-15 (n=76), none underwent placement of an intracranial monitor nor craniotomy and one subject (1.3%) demonstrated a change in his exam (worsened headache and dizziness) concomitant with a progression of his intracranial injury. His symptoms resolved with readmission to the ICU and continued observation. For subjects presenting with ITSAH and a GCS of 15 (n=55), the negative predictive value was 96.3% for no subsequent changes in neurologic exam or symptoms and 100% for a discharge GCS of 15 without an intervention.

Patients with ITSAH have less severe brain injuries than other TBI patients. ITSAH patients with presenting GCS scores of 13-15 demonstrate low rates of clinically significant progression, and when progression occurs it resolves without further intervention. This subset of TBI patients does not appear to benefit from ICU admission.

The EXACT (EXAmen Cognitif Abrégé en Traumatologie) is a new test specifically design to briefly assess all cognitive functions of TBI patients in order to determine orientation at the time of discharge. This instrument is composed of 22 subtests (maximum score totaling 100 points) which can be performed in 15 to 45 minutes. The aim of this study was to verify if the EXACT discriminates patients according to the severity of the TBI and to examine the influence of certain variables on their short-and long-term outcome.

The EXACT was administered to 194 patients who sustained a TBI (77 mild, 61 moderate and 56 severe) at the time of the hospital discharge, to 90 normal control participants and to12 patients hospitalized for orthopedic injuries only (age range from16 to 96 years). The Disability Rating Scale (DRS) and the Glasgow Outcome Scale Extended (GOSE) were administered one year after the accident.

Analyses showed that the total score on the EXACT varies according to the severity of the TBI. However, the performance of the patients hospitalized for orthopedic injuries were equivalent to those of mild TBI patients. Moreover, the EXACT scores were negatively associated with age, the Glasgow coma scale, therapeutic intensity level and the length of stay in intensive care unit. There was also a significant correlation between the score on the EXACT and the functional outcome of the patients one year after their accident. Finally, substance abuse affects the EXACT performance in TBI patients over 50 years old.

The EXACT appears to be a useful brief clinical test to assess the global cognitive functioning of patients during the acute phase of TBI and to predict their short-and long-term outcome.

Blunt Cerebrovascular Injuries (BCVI) are more prevalent than originally believed, with many studies reporting up to 1.5% of blunt traumas that result in BCVI. Prior studies have shown that the mechanism of injury likely involves rapid torsion of the head. We evaluated the relationship between high and low impact mechanism of trauma in relation to number and severity of vessel injury.

Retrospective observational study of patients diagnosed with BCVI between October 2003 and April 2013 at a large Level 1 trauma center. All patients were evaluated by CT angiogram of the head and neck. The BCI grading scale was used to assess radiographic vessel injury. Falls were classified as low impact mechanisms of trauma and motor vehicle collisions (MVC), motor pedestrian collisions (MPC), motor cycle collisions (MCC) and ATV accidents were classified as high impact mechanisms of trauma.

There were 302 subjects diagnosed with BCVI which were primarily male (64%). The type of impact (low versus high) had no association with the number of vessels injured (p=0.11) or with the severity of vessel injury (p=0.97). High impact injuries were more likely to result in a carotid injury (Chi-square=5.13, p=0.02) compared to vertebral injury. Prior retrospective studies have shown similar findings in regards to mechanism of injury and its predictive value.

There is no clear association of mechanism of injury with the number of vessels injured and the grade of injury in this study. Further investigation and greater detailed information about the mechanism of trauma is needed to create a model for reliably predicting vessel injury. Additional information in regards to the use of CT angiogram compared to conventional angiogram as a screening method is also needed.

Spinal cord injury initiates a cascade of local and systemic persistent inflammatory processes that lead to secondary injury and impair functional recovery. Tumor necrosis alpha (TNFa) is an important mediator of secondary injury, partly by increasing synaptic localization of calcium-permeable AMPA receptors in the injury penumbra (Ferguson et al., 2008) . Inhibition of TNFa with a sequestering agent, soluble TNF-receptor 1 (sTNFR1), results in a dose-dependent improvement in behavioral outcome in a rodent model of cervical spinal cord injury as well a reduction in peri-lesional microglial activation (Huie et al., in preparation) . The present study investigated the effects of delayed intrathecal administration of sTNFR1 on peripheral inflammatory profiles.

Long-Evans rats (n= 33) received 350 ng sTNFR1 90 minutes following C5 laminectomy and unilateral cervical contusion injury (75 kdyn force) or C5 laminectomy without injury. Serum samples and liver and spleen homogenates were analyzed for expression of a panel of inflammatory markers by Luminex multiplex assay, western blot, and quantitative PCR.

Intrathecal sTNFR1 was found to reduce levels of key pro-inflammatory cytokines in serum as well as expression of proinflammatory genes and downstream mediators in spleen and liver. Intrathecal sTNFR1 promoted splenic and hepatic expression of markers of 'M2' macrophage polarization, which is implicated in CNS repair and limiting secondary inflammation.

These data highlight the importance of peripheral immune responses following spinal cord injury. In addition, improved behavioral outcome following the delayed inhibition of inflammatory signaling in the lesion microenvironment with sTNFR1 results in a concomitant decrease in peripheral pro-inflammatory marker expression and stimulation of 'M2' macrophage polarization.

Generalized convulsive status epilepticus (GCSE) is associated with significant morbidity and mortality and can be a complication in patients with TBI. The objective of this study is to determine if TBI was an independent predictor of mortality in patients hospitalized with GCSE, and to identify prevalent comorbidities.

We queried the 2002-2010 Nationwide Inpatient Sample databases to identify all patients aged ≥18 years with GCSE using ICD-9-CM codes 345.3. Patients with TBI were identified using ICD-9-CM codes 800.0-801.9, 803.0-804.9, 850.0-854.1 and 959.01. Logistic regression was used to determine if TBI was an independent predictor of mortality in patients with GCSE.

From the inpatient-weighted sample, 112435 patient's aged ≥18 years with GCSE were identified and further analysis was done on this cohort. 5547 patients (4.9%) also had TBI. 

Among patients with a primary diagnosis of GCSE, TBI is not an independent predictor of mortality, after correcting for covariates.

Hypertonic saline (HTS) solutions are commonly used to manage increased intracranial pressure (ICP) and cerebral edema in patients with brain injuries. Though HTS is widely considered effective, the side effects of this therapy have not been fully described in the literature. The purpose of this study is twofold: to describe our HTS scale and its effects on a variety of pertinent laboratory values as well as to attempt to more accurately define the potential risks using HTS therapy in patients with brain injury.

The study includes 153 patients admitted to the neurocritical care unit over 21 months. A 3% NaCl sliding scale was administered to patients with brain injuries who were clinically suspected of benefiting from such therapy. Total time on protocol was dictated by clinical course. Patients are identified by age, sex, race, and diagnosis. The primary outcome measures were blood sodium levels, namely the time to goal (140 mEq/L) and percentage of time below goal after reaching target. Secondary outcome measures include defining the incidence of hyperchloremic metabolic acidosis (HCMA), deep vein thrombosis, pulmonary embolism, and acute renal failure while on the protocol.

The average sodium levels ranged from 136 -148. The average length of time on the protocol was 132.6 hours, while the average time to goal was 15.3 hours. After reaching goal, patients fell below goal an average of 10.5% of the total time on the protocol. The average pH low was 7.34. DVTs and PEs developed in two patients. There were no acute kidney injuries associated with administration of the protocol.

Our 3% HTS scale shows promising and effective results on blood sodium levels in patients with brain injuries. Additionally, side effects appear to be minimal.

The ideal level of sedation in the acute phase treatment of severe TBI patients remains uncertain. The purpose of this study was to compare the effects of light versus heavy sedation on cerebral metabolism by utilizing cerebral microdialysis.

We retrospectively analyzed data from 18 patients with severe TBI (GCS≤8) monitored with cerebral microdialysis who were initially treated with light sedation (propofol<60 μg/kg/min or midazolam<10 mg/hr) and then escalated to heavy sedation (midazolam>10 mg/hr or pentobarbital). The cohort was divided into four groups: 1) low dose midazolam transitioned to high dose midazolam (n=4), 2) low dose midazolam transitioned to pentobarbital (n=5), 3) propofol transitioned to high dose midazolam (n=4), and 4) propofol transitioned to pentobarbital (n=5). All parameters were compared in a within subjects analysis.

When transitioned from light sedation (either propofol or midazolam<10 mg/hr) to pentobarbital (Groups 2&4) the mean GCS decreased by 2 points (p<0.05, p<0.005), whereas when transitioned from light sedation to high dose midazolam (Groups 1&3) there was not a significant decrease (p=0.63, p=0.07). The mean arterial pressure, jugular venous oximetry, oxygen saturation, and core temperature did not have statistically significant changes in any of the groups during the transition from light to heavy sedation. The ICP did not change in any of the groups, although the incidence of elevated ICP>20 mmHg did decrease by over 40% in Groups 2&4. Microdialysis measurements of glucose and the lactate/pyruvate ratio did not significantly change in any of the groups.

Escalating treatment from light to heavy sedation was not associated with significant changes in any physiologic parameters or markers of cerebral metabolism in this cohort. A prospective randomized crossover trial with a significantly larger sample size is required to determine if the reduction in elevated ICP seen on switch to pentobarbital is replicable and clinically meaningful.

Metabolic crisis is a state of energetic compromise after TBI which may occur due to cerebral ischemia or non-ischemic alterations in brain metabolism. The purpose of this study was to determine whether the tissue around the microdialysis probe was ischemic by analyzing the apparent diffusion coefficient (ADC) and investigate whether any ischemia seen correlated with metabolic crisis.

Clinical, conventional MRI, and diffusion MRI data were collected on 21 patients with severe TBI who were monitored with cerebral microdialysis. A region-of-interest (ROI) of 1 cm 2 around the microdialysis probe was used to calculate mean ADC values, and a ROI on the contralateral side was drawn to serve as an internal control. Metabolic crisis was defined as the simultaneous elevation of the microdialysate lactate/pyruvate ratio >25, and low glucose <0.8 mmol/L.

The mean ADC in the microdialysis region was 910 ± 182 μm 2 /sec and in the contralateral region was 872 ± 157 μm 2 /sec, both indicating normal non-ischemic white matter. 5/21 patients (24%) had edema in the microdialysis region with corresponding ADC values > 1000 μm 2 /sec. No patients had ADC values that were in the ischemic range. Patients spent on average 27% of the time of microdialysis monitoring in metabolic crisis. 7/21 patients (33%) were never in metabolic crisis, and 6/21 (29%) spent more than 50% of the time of monitoring in metabolic crisis. There was no significant relationship between ADC values and the burden of metabolic crisis (Pearson r = -0.27, robust r = -0.33).

Cerebral metabolic crisis was present in the majority of patients despite ADC values suggestive of normal appearing white matter. This provides support to the theory that metabolic crisis is often non-ischemic, and suggests that microdialysis results can be extrapolated beyond the focal measurements to represent the global metabolism of normal white matter.

The aim of neurocritical care is to prevent secondary brain injuries in order to improve outcome of TBI patients. All TBI patients may not require the same therapeutic interventions to achieve this goal. The TiL and the APACHE score can be used to reflect treatment and disease severity. The goal of our study is to examine the relationship between TiL, APACHE II score and various outcome measures.

We retrospectively calculated daily TiL throughout neurocritical care stay and admission APACHE score of 34 TBI patients (18 moderate and 26 severe) admitted between 2011 and 2013 at a level one trauma centre in Montréal, Canada. 77% were male and had a motor vehicle accident (71%). At hospital discharge, we administered the EXACT, a new test designed to assess the global cognitive functioning of TBI patients. The Exact is related to outcome at one year. We also performed the Glasgow Outcome Scale extended (GOSE) and the Disability Rating Scale (DRS) at one year.

Mean baseline GCS was 8. The average ICU length of stay was 7 days. Average hospital stay was 28 days. Analyses showed a statistically significant correlation between TiL and the EXACT score (r = -.45), while the correlation with the APACHE score was not significant (r = -.18). There was no significant relation between both scores and GOSE or DRS at one year. The EXACT correlates with outcome at one year.

TiL is correlated with cognitive outcome at discharge and seems more specific to TBI than APACHE II score. Whether TiL is a marker of disease severity or it directly affects outcome needs to be determined. More patients are needed to assess correlation at one year. TiL should routinely be included in baseline characteristics of TBI patients when conducting a study in order to compare groups appropriately.

Contralateral lesion emergence post decompressive craniectomy in traumatic brain injury is a complication described in the literature. Association of this lesion formation with factors related to coagulopathy in trauma has not been well documented. Our goal was to define the incidence of new contralateral intracranial lesions post decompressive hemicraniectomy in blunt traumatic brain injury, and explore the association between factors that contribute to coagulopathy.

We retrospectively reviewed the records and imaging of all patients treated with hemicraniectomy for blunt traumatic brain injury at our institution from May 2007 up to and including January 2012.

Twenty patients were identified during the time period to have undergone decompressive craniectomy for blunt head injury. The average age and Glasgow Coma Scale (GCS) on presentation was 44.1 years (range: 19 -72 years) and 6.5 (range: 3 -14) respectively. All but one patient presented with an extra-axial hematoma as their surgical indication for craniectomy. Seven patients (35.0%) developed new contralateral lesions post-craniectomy. The average peri-operative pH, bicarbonate (HCO 3 ) and hematocrit (HCT) levels for those with new contralateral lesions were lower than those without new lesions. Five of the 7 patients (71.4%) with new lesions had abnormalities on their laboratory results that have been know to be attributable to coagulopathy, with 4 (57.1%) having two or more abnormal results. Eight of 13 (61.5%) patients without new lesion had laboratory abnormalites, with 5 (38.5%) having 2 or more abnormalities identified.

The incidence of new contralateral lesions post-craniectomy for blunt head injury is 35.0% in our experience. There is an association between the metabolic derangements linked to trauma related coagulopathy and the formation of new lesions.

Trauma represents the most common cause of cerebrospinal fluid (CSF) leak. CSF leak carries a significant morbidity and mortality if not promptly recognized and treated. The goal of this study is to focus on CSF leak after trauma and methods of management.

Over a period of 2 years, 79 patients presented with CSF leak following head trauma. CT scan was performed in all cases to detect the site of skull base defect.

CSF rhinorrhea was the most common form of CSF leak found in 49 patients followed by CSF otorrhea. Conservative management of CSF leaks was followed in all cases and included; strict bed rest, prophylactic antibiotics, elevation of the head, avoidance of straining, lumbar puncture or drain and results in resolution of the majority of CSF leaks within one week. Surgery was needed in 19 cases in whom conservative treatment didn't give a response. Endoscopic endonasal approach was used in 12 cases and transcranial repair in 7 cases. In two patients, ventriculoperitoneal shunt insertion was required. Out of the 79 cases, 5 patients developed meningitis due to CSF leak

CSF leak is not common after head trauma however it can lead to significant morbidity and mortality. Most cases with CSF leak respond well to conservative treatment. Surgery is required in cases with persistent CSF leak after conservative treatment and should not be delayed. Endoscopic endonasal repair represents the most suitable option for surgical treatment of CSF leak with high degree of success. 

Leukocytosis is associated with hemorrhage volume and early neurological deterioration after intracerebral hemorrhage (ICH). We examined total white blood cell (WBC) count and absolute monocyte (AMC) and neutrophil (ANC) counts as potential readily available prognostic biomarkers in human ICH.

In a retrospective study, adult patients aged ≥ 18 years who presented to one of two local hospitals with nontraumatic ICH from July 2008 to December 2009 within 12 hours of symptom onset were identified. Demographics, Glasgow coma scale (GCS), ICH volume, ICH location, and 30-day case-fatality rates were determined. Total WBC count, ANC, AMC, and hemoglobin concentration were determined. Linear and logistic regression were used to evaluate factors associated with baseline ICH volume (log transformed) and 30-day case-fatality, respectively.

Of the 186 patients, mean (±SD) age was 67.3±14.8 years, 51% were male, 22% were black. Median [IQR] ICH volume was 12.8 [4.9, 29.4] ml. After adjusting for patient age and initial hemoglobin, higher initial WBC count (p=0.0009) and higher ANC (p=0.006) were associated with higher ICH volume, whereas AMC was not (p=0.4). After adjusting for patient age, GCS, intraventricular hemorrhage (+/-), stroke location, and ICH volume, baseline AMC was associated with greater odds of 30-day case-fatality (OR 2.26, 95% CI 1.10-4.65, p=0.03).

The association of AMC with higher 30-day case-fatality after ICH is hypothesis generating. Given the lack of association between presenting AMC and ICH volume, AMC may contribute to secondary injury after ICH (hematoma expansion and/or cerebral edema).

Hypernatremia is associated with increased mortality in critically ill patients as well as those hospitalized in a neurological intensive care unit. In patients with primary intracerebral hemorrhage (ICH) with critical illness it is not known if serum sodium at the time of discharge can predict subsequent risk of all-cause mortality following hospital discharge.

We performed a two center observational study of patients with ICH who were treated in intensive care units between 1997 and 2007. We studied 3,345 critically ill patients diagnosed with ICH who survived hospitalization. ICH was defined as the presence of ICD-9 code 431 during the hospital admission. The exposure of interest was serum sodium within 24 hours of hospital discharge (categorized as Na ≤ 145 mmol/L and Na > 145 mmol/L). The primary outcome was all cause mortality in the 30 days following hospital discharge. Secondary outcomes included 90-day and 365-day mortality following hospital discharge.

In patients with ICH who received critical care and survived hospitalization, the Na at discharge was a robust predictor of all cause mortality and remained so following multivariable adjustment. Patients with a discharge Na > 145 mmol/L have an OR for mortality in the 30 days following hospital discharge of 4.11 (95%CI, 2.77-6.09; P<0.0001) and an adjusted OR of 3.37 (95%CI, 2.20-5.15; P<0.001) relative to patients with a discharge Na ≤ 145 mmol/L. Similar significant robust associations post multivariable adjustments are seen with death by days 90 and 365 post-discharge.

In patients with ICH treated with critical care who survive hospitalization, an elevated Na at discharge is a robust predictor of subsequent all cause patient mortality. Increased Na at discharge may reflect more aggressive hyperosmolar therapy during admission, end organ dysfunction, poor oral intake or a combination thereof which may explain the observed impact on patient survival following discharge.

As life expectancy rises, stroke is expected to affect older individuals. A significant growth in the incidence of intracerebral hemorrhage (ICH) in the elderly occurred over the last decades. However, the very elderly (>80 y) were not adequately studied and data about their clinical characteristics and outcomes are still lacking in the literature.

We studied data of consecutive patients from a prospectively collected hospital stroke database, from august 2008 to april 2013. Clinical characteristics and outcomes of 31 very elderly patients aged 80-97 years with ICH were compared to 100 younger patients with ICH.

Patients exhibited similar gender proportion in the groups. Compared to younger controls, very elderly patients had higher NIHSS score at admission (mean 6.1 vs 10.8, p=0.03), higher ICH scores (median 1 [interquartile range, 0-3] vs. 2.5 [1] [2] [3] [4] ; P=0.002), more history of previous stroke (14.3% vs. 32.1%, p=0.049), higher prevalence of atrial fibrillation/flutter (5.1% vs. 28.6%, p=0.001) and coronary artery disease (4.1% vs. 42.9%, p<0.001). Patients from the very elderly group also had poorer outcomes (mRs 3-6) than the younger group (16.1% x 56.7%, p<0.001). In a logistic regression analysis, ICH score was found to be a predictor of bad outcome only for younger patients. (OR 2.7, CI 1.52-4.9 for mRs 3-6, p=0.001).

Patients aged 80 years or more with ICH do not share the same clinical features as their younger counterparts. In addition to a higher prevalence of risk factors (previous stroke, atrial fibrillation/fluter and coronary artery disease), they exhibit higher NIHSS score at admission and worse outcomes. Unlike with younger patients, the ICH score did not predict functional outcome in the very elderly. One hypothesis is that age is a stronger prognostic indicator in this group. Better outcome prediction markers should be investigated in future studies with attention to the very elderly age group.

Financial Support: None

The incidence of generalized convulsive status epilepticus (GCSE) following intracranial hemorrhages (ICH) has been studied only in small patient populations. Given the different pathophysiology, we hypothesized that the incidence of GCSE will vary with the type of ICH.

We queried the 2002-2010 Nationwide Inpatient Sample databases to identify all patients aged ≥18 years with traumatic and non traumatic intracranial hemorrhage (ICH); subarachnoid (SAH), subdural (SDH), epidural (EDH), and intraparenchymal hemorrhage (IPH) using the ICD-9-CM codes 852.x, 430, 432.1, 432.0 and 431 respectively. Patients with GCSE were identified using ICD-9-CM code 345.3. Logistic regression was used to compare the incidence of GCSE in each ICH type.

From 2002-2010, among 1,589,465 patients aged ≥18 years that had intracranial hemorrhage; 586,673 (36.9%) had IPH, 579,441 (36.5%) had SDH, 408,130 (25.7%) patients had SAH, and 15,221 (0.9 %) patients had EDH. Patients with SDH had higher prevalence of congestive heart failure, coagulopathy, hematoma drainage; those with IPH had higher rates of hypertension, mechanical ventilation, those with SAH had higher rates of hyponatremia, ventriculostomy and brain infections whereas those with EDH had highest rates of craniotomy/craniectomy. The incidence of GCSE was highest for SDH (0.4%), followed by IPH (0.3%) and 0.2% for both SAH and EDH. When adjusted for confounding variables, as compared to EDH, the odds of GCSE were higher for SDH (OR 2.44, CI 1.48-4.02) and IPH (OR 1.8, CI 1.09-2.98). For GCSE, the mortality was highest for IPH (28.5%) followed by SAH (17.1%), SDH (12.2%) and EDH (7.1%).

Incidence of GCSE is highest in patients with SDH (two fold higher than that of SAH and EDH and 50% higher when compared to IPH). This is an interesting finding as in these individuals blood is not in direct contact with the brain parenchyma.

Financial Support: None

The incidence and significance of thrombocytopenia among patients with intracranial hemorrhage (ICH) are not well established. We performed this retrospective chart review to gain more insight into the characteristics of ICH in this subset of patients.

We screened the charts of all patients at our institution who were diagnosed with ICH between January 2009 and March 2012. Thrombocytopenia was defined as a platelet count of <100,000/mm 3 on admission. We included in the study only patients with primary, i.e. unprovoked ICH. Thus, patients with hemorrhage into an ischemic stroke or mass lesion, arterio-venous malformations and trauma were excluded. We reviewed all available laboratory and clinical information, as well as imaging studies of the patients who were included in the study.

Among 487 consecutive ICH patients, 31 (6.4%) were thrombocytopenic. Following exclusion of seven patients with secondary ICH, 24 patients were included in the study. Ten patients (42%) were women and mean age was 62+/-16 years. On admission, mean Glasgow Coma Scale was 11 and mean ICH score was 1.7. Five patients had <50,000 platelets/mm 3 . Sixteen patients (67%) had a known history of hypertension, but only 4 patients (17%) had a recorded blood pressure in excess of 140/80 mmHg in the 24 hours preceding the diagnosis of ICH. Ten patients (42%) carried a diagnosis of liver cirrhosis. Eleven patients (46%) had lobar hemorrhage, in 2 cases multifocal. Intraventricular hemorrhage (IVH) was found in 10 patients (42%). In-hospital mortality was 29%, whereas 17% were discharged to a skilled-nursing facility and 12.5% to a long-term acute care facility.

In this retrospective review we found that patients with thrombocytopenia and ICH were not likely to have hypertension at the time of onset. ICH was frequently lobar and associated with IVH. In-hospital mortality rates were similar to those reported in the general ICH population.

Warfarin-associated intracranial hemorrhage (wICH) can be a devastating outcome of therapy. Few studies have examined prothrombin complex concentrate (PCC) versus recombinant factor VIIa (rFVIIa) in the treatment of wICH. The purpose of this study is to compare the safety and efficacy 3-factor PCC versus rFVIIa in the treatment of wICH.

A retrospective chart review was conducted for eligible patients with wICH that received either PCC or rFVIIa and were admitted between January 2008 and November 2012.The primary objective was time to INR normalization. Secondary endpoints included the degree of hematoma expansion, intraoperative hemostasis, and incidence of thromboembolic complications.

The cumulative incidence of INR reversal at 120 hours post-infusion for PCC (n=20) and rFVIIa (n=25) was 95% and 93%, respectively. PCC resulted in a faster median time to normalization (5 vs. 29 h, p = 0.29). All subjects who received PCC preoperatively and 71% who received rFVIIa achieved adequate intraoperative hemostasis (p = 0.13). Thromboembolic complications occurred in two patients in each group (p > 0.99). Patients in the rFVIIa group had a larger median hematoma expansion score at baseline (81 vs. 40, p = 0.14) and the degree of hematoma reduction was greater in this group (-65 vs. -23, p = 0.57). The median score at discharge was 1.3 and 21 in the PCC and rFVIIa groups, respectively.

There was no difference in safety or efficacy between PCC and rFVIIa in the treatment of wICH. However, PCC may provide a more rapid achievement of INR reversal and result in more adequate intraoperative hemostasis.

Emerging data find decreased mortality and improved functional outcome in female mice in experimental intracerebral hemorrhage (ICH) models compared to male mice. We postulate that sex-specific outcomes may occur due to differences in sex chromosome gene expression.

Ten to twelve week old male and high estrous female C57/Bl6 mice (n=3/group) underwent ICH induction via left intrastriatal collagenase injection. Whole brain samples were collected at baseline, 2, 6, 24, and 120 hours after injury. Genome-wide expression profiling was performed with Affymetrix GeneChip Mouse Genome 2.0 to identify genes differentially expressed between baseline and each timepoint in males and females. Comparisons were made between baseline and each post-injury timepoint to determine significant expression of genes located on the X-chromosome in both sexes and the Y-chromosome in males. Also, the ratio of normalized expression in females versus males of each Xchromosome was assessed for each timepoint to identify genes differentially regulated between sexes.

No differential gene expression was observed for male X-or Y-chromosomes post-ICH. In females, GM5124 and SEPT6 were downregulated at 24 hours post-ICH, whereas LAS1L and TIMP1 were upregulated at 24 and 120 hours post-ICH. Direct male-to-female comparisons at each timepoint revealed up-regulation of ALAS2of 2.19 times higher at 2 hours and TIMP1 of 2.91 times higher at 120 hours post-ICH in females compared to males.

Differential gene expression occurs on the female X-chromosome in mice after experimental ICH. While GM5124 is a mouse gene, future investigation of LAS1L (role: ribosomal biogenesis), TIMP1 (role: tissue inhibitor of metalloproteinases), SEPT6 (role: cytokinesis) ALAS2 (role: heme biosynthetic pathway catalyst), and their associations with sex-specific outcome after experimental ICH may yield novel therapeutic targets for translation into the human disease.

Emerging data suggest decreased mortality and improved functional outcome in female mice after intracerebral hemorrhage (ICH) compared to male mice. Improved function in female mice is associated with smaller hematoma volume and less cerebral edema compared to male mice. We postulate that sex-specific differential hemostatic pathway expression may shed mechanistic light on these findings

Ten to twelve week old male and high estrous female C57/Bl6 mice (n=3/group) underwent ICH induction via left intrastriatal collagenase injection. Whole brain samples were collected at baseline, 2, 6, 24, and 120 hours after injury. Genome-wide expression profiling was performed with Affymetrix GeneChip Mouse Genome 2.0 to identify genes differentially expressed between baseline and each timepoint in males and females. Pathway analysis was then conducted to identify hemostatic genes.

There were no differences in hemostatic genes expressed between male and female mice at baseline or 5 days post-ICH. Compared to baseline, hemostatic genes were uniquely expressed at 2 hours (3 in male and 13 in female), 6 hours (2 in male and 22 in female), and 24 hours (2 in male and 7 in female) post-ICH. Differential expression of Serpin-1 and -1b (for males), PAI-1, ADAMTS13, CD36, CD46, VWF, STK16, and VEGFA (for females) occurred consistently at each timepoint after ICH.

Sex-specific hemostatic gene expression occurs after experimental ICH. Differential expression of these genes may play a role in determining hematoma volume, edema, and, ultimately, neurological function after ICH. Future investigation should focus on the function of these genes after ICH and their potential for translational as therapeutic targets.

Spontaneous Subdural Hematoma's (SDH) have been reported in weight lifter's taking androgens. We report a case of a male weight lifter taking testosterone acquiring a spontaneous SDH. The SDH resulted in a seizure and four thoracic vertebral fractures due to his remarkable muscle mass.

Spontaneous SDH is most frequently seen in elderly patients who are on anticoagulants or who have a coagulopathy due to thrombocytopenia or liver disease. We report a case of a 47 year old healthy man with no comorbidities who weighed 165 pounds but was bench-pressing 250 pounds twenty times each afternoon. He developed a headache after one such session so went to bed. When he awoke he had dysarthria and back pain. His family drove him to the emergency department later in the day. The workup revealed a 6 mm right sided frontal-parietal SDH with 3 mm of right to left midline shift. An MRI of the spine showed thoracic vertebral compression fractures at T5, T6, T7, and T9. He admitted to noting he had wet his pants when he woke in the morning and on exam he had bitten his tongue. Although no one witnessed the seizure, it must have occurred while he was in bed, and since he was post-ictal he had no memory of it. He did have some rhabdomyolyis with CPK's of up to 6000 and was hydrated aggressively and he did not develop renal failure. He did not require surgery for his SDH and was discharged on anticonvulsants and with a clam shell vest. .

Spontaneous SDH is being reported more frequently in healthy weight lifters taking androgens or testosterone. Further investigation of this association of male hormones and high venous pressures resulting in subdural bleeding is indicated

Intracerebral Hemorrhage is a medical emergency, characterized by high morbidity and mortality, which should be promptly diagnosed and aggressively managed. Initial monitoring and management of ICH patients should take place in an intensive care unit with neuroscience expertise. Hemorrhagic Stroke Patients coming to the ED were not being treated consistently in the areas of assessment, documentation, blood pressure control, and INR reversal and affected the time it takes to transfer to Neuro Critical Care.

A team was formed that included Quality Outcomes, Emergency Department, Neuro Critical Care, and Neurosurgery. This team reviewed all hemorrhage patients weekly and found inconsistencies in written orders, BP medications utilized, documentation of vital signs and neuro checks. The Stroke Alert Process was then redesigned to include Hemorrhagic strokes. An overhead page occurs and a Neurosurgery representative arrives at bedside to assess and write orders. Brief Intracranial Hemorrhage orders were developed to include INR reversal, specific BP parameters, and medications. An ED Nurse champion has been assigned to assist with real-time auditing of vital signs and neuro checks, documentation and coordination of care. The neuro documentation system in the ED was also revised to include a more thorough assessment.

Door to BP control time has decreased to 35 mins, Pro-coagulant given time has decreased to 23 mins, Bed assigned to arrival in NCC has decreased to 67 mins, Documentation compliance has improved from 67% to 99% for completion of vital signs and neuro checks every 15 minutes.

The team continues to meet weekly to review all hemorrhagic stroke patients admitted thru the ED. Concurrent Auditing is done daily on new admissions. The process also includes BP management medications being premixed and available 24/ 7 in the ED PYXIS. A reminder label is placed by Pharmacy on the Pro-coagulant medication to redraw INR within one hour.

There have been a number of reports in the literature correlating improved outcome with patient volume and there have been a few reports demonstrating improved outcome in hospitals with Neurocritical care units. The idea of triage to centers of excellence is based on these types of results. This report will examine these correlations from a large multicenter national database accrued from the ERICH (Ethnic/Racial Variations of Intracerebral Hemorrhage) study. :

This examination of the ERICH database will be simply a descriptive display of mortality and outcome per hospital (deidentified) lined up by size and then grouped by presence of a neurocritical care unit in the hospital. Currently, there are 41 hospitals that have enrolled 1872 ICH patients into the ERICH database. These hospitals have a large range of different characteristics including overall size, number of ICU beds and presence or absence of neurocritical care units.

Results will be presented at the NCS meeting to maximize the number of patients examined (recruitment is still ongoing).

It will be expected that the larger hospitals with neurocritical care units will have the best outcomes.

The epidemiology of circulatory shock after spontaneous intracerebral hemorrhage (ICH) is unknown. We sought to determine its incidence, risk factors, and effect on case-fatality. We tested the hypothesis that shock was not associated with higher case-fatality.

Retrospective multi-center cohort study of 83 ICUs in the United States from 2003-2008. ICH patients >17 years of age admitted to an ICU. Shock was defined as sustained systolic blood pressure <90mmHg for ≥1-hour despite vasopressors.

A total of 4,192 patients. Median age was 67yrs(IQR 54-77), 2221(53%) were female, and 3030(75%) were White. Median APACHE-II score was 15(interquartile-range [IQR] 11-21) and GCS was 11(IQR 6-14). Shock was present in 212(5%) patients. Case-fatality was 72% among shock vs. 30% without shock (p<.0001). In the cohort, 18 patients had pulmonary artery catheters (PA-C) placed. The PA occlusion pressure in patients with shock was 10mmHg(IQR 5-13) vs. 16mmHg(IQR 8-18) without shock (p=.05), cardiac index was 3L/min/m2(IQR 2.5-3.8) in shock versus 3.8L/min/m2(IQR 2.8-4.4) without shock (p=.3), CVP was 5mmHg(IQR 2-11) in shock versus 11mmHg(IQR 8-15) without shock (p=.08), and PA systolic were 26mmHg(IQR 18-30) in shock versus 37mmHg(IQR 30-47) without shock (p=.03). In multi-variable analysis the following were associated with increased case-fatality: age (OR 1. , and shock (OR 1.9, 95% CI 1.2-3.0). EVD placement was associated with survival (OR 0.8, 95%CI 0.6-0.9).

In ICH patients admitted to an ICU, there was a 5% incidence of circulatory shock. In those with a PA-C, the hemodynamic profile suggested hypovolemic shock, which was associated with increased case-fatality. Our findings require further study.

Background : Nonconvulsive seizures (NCs) are frequent after intracerebral hemorrhage (ICH) and associated with mass-effect and poor outcome.

Objective: To study NCSz on surface and intracortical depth EEG in ICH patients. Methodology: We prospectively studied 17 consecutive adults (57+/-12 years-old, 53%[N=9/17] female) with non-traumatic ICH that underwent invasive multimodality brain monitoring (MMM) including minidepth EEG between 6/2006 and 1/2013. We assessed the predominant EEG background and identified each minute with seizures surface or depth EEG recordings. Seizures lasting at least 5 minutes and preceded by 30minutes without any seizure activity were defined as new-onset seizures.

Underlying ICH etiology was most frequently hypertension (59%;N=10) located most frequently in the basal ganglia (53%;N=9) or cortex (18%;N=3). 35%(N=6) underwent surgical evacuation and 12%(N=2) had hemicraniectomy. MMM probes were placed ipsilateral to the ICH in 60% (N=9), median monitoring length was 8.5 days (IQR 5.5,13.5). Seizures were identified in 35% of patients (6% surface and depth, 29% depth only; 2086 minutes of cumulative intracortical and 161 minutes surface seizure-duration) amongst 77 days of EEG recording. 19 new-onset seizures were identified in 4 patients (median duration 13 [IQR8,18] minutes; 47%[N=9] lasting ≥15 minutes). Presence of seizures was not associated with demographics, ICH score, surgical interventions, or ICH location. During monitoring no positive CSF cultures were observed and no bleeding was identified after placement or removal of the probes. Modified Rankin score(mRS) at 3 month was 5(IQR4,5; 71% severely disabled or dead). No relationship between intracortical seizure, surface seizure or background attenuation and 3 months outcome was identified.

In this preliminary cohort of ICH patients with intracranial EEG monitoring we confirmed a high frequency of depth seizures but were underpowered to demonstrate an association with outcome. We are currently analyzing the MMM data associated with NCSz in this cohort.

Cerebral blood flow autoregulation (CBFa) is the intrinsic ability of the cerebral vasculature to maintain a stable cerebral blood flow over a wide range of blood pressures (BP). Hematoma expansion and brain perfusion may be affected by acute BP changes, leading to arbitrary BP targets for groups of ICH patients. We used non-invasive continuous indices of autoregulation to test the hypothesis of impaired CBFa in ICH.

Four comatose patients with ICH had continuous near-infrared spectroscopy (NIRS) and transcranial doppler (TCD) monitoring of middle cerebral artery blood flow velocity for three days. The mean velocity index (Mx) was calculated as a moving, linear correlation coefficient between slow waves of middle cerebral artery blood flow velocity and mean arterial blood pressure (MAP). The cerebral oximetry index (COx) was calculated as a similar coefficient between slow waves of cerebral oximetry and MAP. When cerebral blood flow is autoregulated, Mx and COx vary around zero. Loss of CBFa results in progressively more positive Mx and COx.

CBFa was impaired in the ipsilateral side of the lesion in all cases. When there was significant midline shift of ≥10 mm at the level of the pineal, patients had bilateral impaired CBFa. The mean GCS at coma onset was 5 with a mean ICH score of 3. The location of ICH was: basal ganglia and thalamus (50%), midbrain and thalamus(25%), and centrum ovale (25%). The mean ICH volume was 46 cc. All patients had a poor outcome (mRS >2) at the time of discharge.

Our pilot study shows the dynamic changes in CBFa, initially with impairment in the ipsilateral side of the lesion and progression to bilateral impairment with severe mass effect and midline shift. This highlight the need for real-time CBFa monitoring to guide the dynamic BP management based on the pathological evolution of the brain.

A recent meta-analysis reported the use of HMG-CoA reductase inhibitors (Statins) is not associated with increased risk of intracerebral hemorrhage (ICH). In routine laboratory studies, low density lipoprotein (LDL) level is used to predict the effectiveness of statins in lowering cholesterol. The Adult Treatment Panel guidelines (ATP III), recommend a goal LDL of <70 in patients who are at high risk for ischemic heart disease or stroke. To our knowledge, no prospective study has determined whether lower LDL levels secondary to statin use has the same risk as primary hypocholesterolemia in predicting the admission volume and/or early hematoma growth.

This study analyzes data from the national Ethnic/Racial Variations of Intracerebral Hemorrhage(ERICH) data set to explore the association of statin use and low LDL on hematoma volume and early hematoma expansion. Subjects with ICH and at least one admission neuroimaging were included. Univariate analysis is used to examine the association between hematoma volume/expansion and statin use with/without low LDL levels on admission (defined as <70mg/dl).

In the ERICH dataset, 913 cases of spontaneous ICH with admission CT scans were enrolled prior to January 2013. Of these, 575/913 (63%) had LDL values recorded, 155/913 (17%) were on a statin prior to admission. Among statin users 65/155(42%) did not have admission LDL values. Among those with a recorded LDL level, 108/575 (19%) reported LDL have values <70mg/dl on admission. Repeat imaging is available for 570/913 (62%) of case and will be examined to explore hematoma expansion in the setting of statin use[92/570 (16%) cases]; and with LDL as a continuous variable.

Final analyses are pending and will be reported at the annual meeting of the Neurocritical Care Society

Outcome from spontaneous intracerebral hemorrhage (sICH) may depend on variations in patient care and settings. We developed a sICH specific therapy intensity level (TIL) metric based on the performance of evidence-based elements in a cohort of sICH patients.

Cohort study of 170 patients with sICH treated in 2 academic neuroICUs. sICH severity was determined by the ICH score.

Pre-defined quality indicators were identified based on current guidelines, scientific evidence, and ease of acquisition within the first 72 hours of hospital admission. We assessed performance on each indicator and association with discharge mortality. Significant indicators were used to develop a TIL score.

Median ICH score was 3 (iqr 1); discharge mortality was 51.2%. Half of patients (54.1%) were transferred from an outside facility. Seven variables from 20 tested were significantly associated with decreased discharge mortality: no DNR/withdrawal of care status within 24 hours of arrival, effective treatment of acute hypertensive response within 2.5 hours, reversal of INR (<1.4) within 2 hours of first elevated INR, ≥2 reversal agents administered within 2 hours of first elevated INR, target glucose within 4 hours of high glucose detection, no recurrent hyperglycemia within 72 hours of admission, and no recurrent hyperpyrexia within 72 hours of admission. One point was given for each or if not applicable. Controlling for Glasgow coma scale, age and sICH volume, higher 7-point TIL score was independently associated with decreased mortality (OR 0.58; 95%CI [0.40-0.84]; p=0.004). In contrast, shorter time to CT scan acquisition, NCCU admission, starting osmotic therapy and early tracheostomy were associated with increased mortality.

A simplified TIL score based on evidenced-based patient-care parameters within the first three days of admission after sICH can predict early mortality after severity adjustment. The next step is validation of the quality indicators and thresholds on a large series of sICH patients.

Financial Support: Daniel Hanley, MD is funded by CLEAR III 5U01-NS062851-03, and MISITE II5R01-NS046309-07.

The long term resolution of spontaneous intercerebral hemorrhage (sICH) on neuroimaging has not been studied. We hypothesized that the volume of residual hypodensity following sICH is associated with original ICH volume and long term neurologic outcomes.

This was a post hoc exploratory analysis of computed tomography (CT) brain scans collected from three trials evaluating acute therapies for sICH (CLEAR IVH, ICES, MISTIE II). A computerized semi-automated volumetric analysis of CT images within 72 hours of ICH onset, at 30 days, and at 180 days was performed on acute ICH volume and the residual hypodensity (RH) in the absence of blood. RH volumes were compared with early ICH volumes and prospectively collected 30 and 180 day modified Rankin scores (mRS).

Of 129 patients with CT at 30 and/or 180 days, median diagnostic/72 hour ICH volumes decreased significantly from 30.6 (iqr33.2)/36.1 (28.7)cc to RH volumes of 7.7(9.9) and 9.0(13.0)cc at 30 and 180 days respectively. Diagnostic ICH volume was highly correlated with RH volume at 30 and 180 days (p<0.001). RH was most commonly located central to initial ICH midpoint (79.7%). Median RH volumes at 30 and 180 days were significantly associated with poor outcome (mRS 4-6) at both time points (p=0.04, p=0.03, respectively). Median RH volume decreased by 0.72cc in patients with good outcome at 180 days (mRS 0-3) and increased by 4.43cc in patients with poor outcome (p=0.02). Only diagnostic ICH volume was significantly associated with >5cc increase in RH volume between day 30 and 180.

Residual hypodense regions at 1 and 6 months post sICH are significantly smaller than initial hematoma size and may be associated with long term neurologic outcomes. Increase in RH volume over time may contribute to a poor recovery. The significance of these regions to hematoma formation and regression warrants further investigation.

Financial Support: Dr Daniel F. Hanley was awarded significant research support of grants number R01Ns046309 and 5U01NS062851.

Acute intensive blood pressure (BP) reduction may improve outcomes in patients with spontaneous intracranial hemorrhage (ICH). However, there is concern that intensive BP lowering will lead to acute kidney injury (AKI). The purpose of this study was to evaluate the effects of intensive BP lowering on renal function in ICH patients.

A retrospective analysis was conducted in ICH patients from 07/2008 to 06/2012 at a tertiary care facility. Rates of AKI was compared among 3 groups: those who received aggressive systolic blood pressure (SBP) lowering ≥ 25% from baseline (AG), moderate SBP lowering <25% from baseline (MG), and patients who had an increase in SBP (IG) from baseline at 6 hours. AKI was defined by Acute Kidney Injury Network (AKIN) criteria at 48 and 72 hours. Outcomes were assessed using modified Rankin score at discharge and 90 days.

A total of 46, 83, and 87 patients in the IG, MG, and AG groups were included. Mean ΔSBP at 6 hours was an increase in SBP by 13±10mmHg in IG, a decrease by 23±24mmHg in MG, and a decrease by 84±29 mmHg in AG (p<0.01). Baseline mean serum creatinine, median APACHEII and median NIHSS were not different between groups. The incidence of AKI was not different at 48 hours (IG; 26%, MG; 40%, AG; 35%; p=0.65) or at 72 hours (IG; 25%, MG; 28%, AG; 16%; p=0.16) between groups. No difference in mortality, or clinical outcomes was found between the groups. In multivariable regression analysis controlling for baseline characteristics and hemorrhage etiology, change in SBP in initial 6 hours did not predict occurrence of AKI.

Compared to increased BP or moderate BP control, ICH patients who undergo aggressive BP reduction within the first 6 hours do not appear to be at greater risk of AKI at either 48 or 72 hours.

A new measure of determining the lesion load of a canonical CST derived from healthy controls on diffusion tensor imaging (DTI) MRI had excellent predictive value in ischemic stroke. We investigated the utility of the DTI-derived data obtained during the acute phase of ICH, and compared with ICH score in predicting functional recovery.

Retrospectively review 32 patients with supratentorial ICH underwent DTI within 4 days after ictus and with CST affected. We compared the fractional anisotropy (FA) values in 5 slices below the level of the lesion on the affected and unaffected CST; and in the cerebral peduncles (CPs), separately. We calculated the ratio of FA (rFA=FA affected side /FA unaffected side ), and examined the value of the ICH score and rFA in predicting functional outcome assessed by modified Ranking Scale (mRS) at follow-up, using ROC analyses.

The adjusted mean FA values did not significantly differ between the lesion and the non-lesion side (0.54±0.1 vs. 0.54±0.9, p=0.20 for the 5 slices below the level of the lesion; and 0.56±0.7 vs. 0.57±0.5, p=0.35 in the CPs). The rFA at the CPs level, but not slices adjacent to the lesion, were significantly lower in the group with unfavorable outcome than those with favorable outcome (0.96±0.14 vs. 0.99±0.08, p= 0.025). The ICH score at admission had the greatest areas under ROC curve (AUC 0.72-0.84) in predicting functional outcome, vs. AUC 0.69-0.82 when combined with mean rFA, and AUC 0.40-0.51 for mean rFA alone.

In this pilot study, the prognostic value of the ICH score surpassed that of DTI during the acute phase of ICH. Incorporating CST-FA data below the level of the lesion did not improve the predictive value of the ICH score. Larger prospective studies are needed to assess the prognostic role of various DTI-derived measures at different times following ICH. 

Background and Purpose -The purpose of this study was to determine if there is an association between distance from interventional stroke center as well as time to revascularization and functional outcome in patients with acute ischemic stroke undergoing thrombectomy.

Methods -One hundred twenty eight consecutive patients that underwent attempted intra-arterial (IA) revascularization for acute ischemic stroke in 2008-2011 were retrospectively reviewed. Patients were treated with IV-tPA if eligible and selected for IA revascularization based on CT perfusion and clinical findings. Independent variables included in analysis were age, gender, stroke location (anterior vs. posterior), NIHSS, use of IV t-PA, use of IA t-PA, distance from interventional stroke center, time to thrombectomy from onset of symptoms, and TIMI score. Clinical outcomes were determined by a 90-day modified Rankin Score (mRS). Logistic regression analysis determined factors associated with a favorable outcome (mRS≤2).

Results -The mean age was 66 years, and the mean NIHSS was 17 (SDV 6.7). Fifty-one (39.5%) patients received IV tPA, and 111 (86%) of the patients had anterior circulation strokes. Eighty one (62.8%) patients presented to a hospital prior to transfer to the Comprehensive Stroke Center (CSC). After controlling for patient age, gender, NIHSS, and IA-tPA status, initial presentation to a CSC (OR 1.55 95% CI 0.69-3.5), proximity to interventional stroke center (<25 miles) (OR 0.88 95% CI 0.40-1.9) and time to procedure <8 hours (OR 1.03 95% CI 0.47-2.3) were not independently associated with improved functional outcome.

Conclusion -In this single center retrospective study, among patients who underwent revascularization based on CT perfusion imaging, initial presentation to a stroke center, proximity to the interventional stroke center, and time to procedure were not associated with improved functional outcome. These findings suggest that there may be factors other than time that influence outcomes.

Cerebral arteriopathies complicate the clinical course of couple diseases such as acute bacterial meningitis, systemic administration of dihydroergotamine for status migrainosus, and in patients on serotonin selective reuptake inhibitors (SSRI) for depression.

We are describing a case series of four patients with cerebral arteriopathies secondary to bacterial meningitis and induced by dihydroergotamine and SSRI administration who underwent endovacular therapy for arteriopathy.

We are reporting 4 patients two of which presented with streptococcus bacterial meningitis, first patient cerebral angiogram showed severe arterial narrowing in the supraclinoid internal carotid arteries and the basilar artery, patient had Mild to moderate improvement in each vessel caliber after infusion of 20 mg of intra-arterial verapamil. Second patient with severe arterial narrowing of bilateral supraclinoid internal carotid arteries and in the middle cerebral artery distribution bilaterally, infusion of 5 mg of intra-arterial nicardipine into the bilateral internal carotid arteries resulted in significant improvement in the flow of bilateral anterior circulations. Third patient received intravenous dihydroergotamine to abort her migraine attack and developed encephalopathy and focal neurological exam, cerebral angiogram showed severe segmental arterial narrowing in all vascular distributions and infusion of intra-arterial nicardipine resulted in moderate improvement in the diffuse narrowing. Fourth patient had been administered high doses of citalopram for her depression and presented with dizziness, headaches, and encephalopathy. Cerebral angiogram showed bilateral supraclinoid internal carotid mild narrowing with moderate basilar artery narrowing. Intra-arterial nicardipine infusion resulted in slight improvement in the arteriopathy.

Cerebral arteriopathies secondary to meningitis is not uncommon. In addition, arteriopathies described in clinical settings secondary to dihydroergotamine or SSRI administration need to be treated aggressively before developing cerebral ischemia and infarction. Endovascular treatment for refractory cases of secondary cerebral arteriopathy may have some positive effect on this disease course but still need further research and investigation.

Financial Support: None

The benefits of intra-arterial therapy (IA) in patients with large artery stroke (LAS) remain unclear. We compared clinical outcomes at hospital discharge in LAS patients treated with and without IA therapy to determine safety and efficacy of the intervention.

We conducted a case-control study on prospectively collected data. Consecutive LAS patients admitted to a single tertiary care center between July 2006 and November 2011 were analyzed. LAS are defined as NIHSS ≥ 10. Patients who were treated with IA (cases) were compared with patients with persistent major branch occlusion who did not undergo IA (controls). Clinical outcome was assessed at discharge from Get with the Guidelines Stroke-database; functional measures at discharge (independent, requiring assistance, dependent, or deceased).

There were 287 patients with LAS in the study period. Of these, 86 patients were included in the analysis: 39 cases and 47 controls. Baseline characteristic were not significantly different between cases and controls. IA treated patients were more likely to be independent at discharge (25.6% vs. 6.4%; P=0.01), but also had higher in-hospital mortality (17.9% versus 8.5%; P=0.01). The rest of the LAS patients were excluded from study because of symptom resolution after IV t-PA or spontaneously, contraindication for intervention or had no large vessel occlusion on CT/MRI angiography.

Patients treated with IA had a higher rate of independence at discharge, but also higher inpatient mortality than LAS patients who did not undergo IA. Careful patient selection is important in selecting patient for IA interventions as our data showed higher inpatient mortality with treatment.

Recently the Trevo Stent Retriever device was approved for thrombectomy in acute ischemic stroke (AIS) patients≤ 8hrs of symptom onset, but few data are available regarding utility at >8hrs. The aim of this study was to evaluate the safety and efficacy of this device in patients >8hrs.

Retrospective review of all AIS patients evaluated at a large comprehensive stroke center between October 2012-May 2013 in which the Trevo device was used. CTP was routinely used to select patients for treatment. Use of concurrent IV rt-PA or other endovascular procedures, 24hr follow-up CT or MRI, and admit/discharge NIHSS and mRS were reviewed. Safety was assessed by sICH rate. Clinical outcome was determined by NIHSS and mRS at discharge.

45 patients (39%male, mean age 61) underwent revascularization. 25 (56%) received full-dose IV rt-PA prior to IA therapy; 15 had additional IA therapies (n=10; IA rt-PA, n= 7; other mechanical thrombectomy). 41 (91.1%) experienced successful recanalization (TICI 2b or 3). 24 patients were treated ≤ 8hrs and 21 >8hrs. Of 45 patients, 39 had safety and outcome data. Of patients treated ≤8hrs, the mean admit and discharge NIHSS were 11 and 6, and for those treated >8hrs, 16 and 9 (p= 0.54 and 0.44 respectively). Of patients treated ≤ 8hrs, 41% (9/22) achieved mRS ≤ 2 compared with 29.4% (5/17) of those treated ≥ 8hrs (p=0.46). 2 (5%) patients experienced sICH, one in each group. Mortality was 5% (n=1 per group).

Treatment of ischemic stroke patients with the Trevo device ≥ 8hrs after symptom onset appears to be safe, and is associated with a high recanalization rate. Functional outcomes are similar to those treated ≤8hrs after symptom onset.

Financial Support: Dr. English served as a consultant for stryKer Neurovascular.Dr. Barazangi served as a member of speaker's Bureau for Genentech for 2012.

Worldwide, 10 million people survive stroke each year and many subsequently develop dementia. After excluding other risk factors, stroke itself approximately doubles the risk of later developing dementia. We investigated whether neuroinflammation persists in the brain for months after stroke, and if chronic inflammation caused by a stroke can lead to cognitive impairment.

We utilized a mouse model of stroke that generates a large cortical lesion and no immediate cognitive deficits to characterize inflammatory responses (n=7-10 mice per group), long-term potentiation (LTP) (n=6-9), and cognition (n=10) before and after stroke in wildtype C57BL/6J and B-lymphocyte deficient (muMT) mice. Controls underwent sham surgery. All experiments and analyses were performed in a blinded fashion.

The inflammatory response to stroke continued for at least 3 months in wildtype mice, and was characterized by the presence of macrophages and B-and T-lymphocytes in the stroke lesion, and the accumulation of immunoglobulin G (IgG) in the surrounding tissue. Although hippocampal LTP is normal 1 week after stroke, by 7 weeks wildtype mice develop an approximately 50% decrease (P<0.05). IgG also accumulates in the hippocampus, and mice develop behavioral impairment on both Y maze and novel placement testing, between 1 and 7 weeks after stroke. LTP deficits continue to worsen and LTP is nearly absent 3 months after stroke (P<0.05). Cognitive impairment and LTP deficits are mitigated in muMT mice that lack mature B-lymphocytes, and LTP deficits can be reproduced by applying IgG purified from 7-week stroke lesions to acute slices from uninjured mice.

This suggests that stroke stimulates a long-lasting autoimmune response wherein B-lymphocytes manufacture brainreactive autoantibodies that over time directly impair brain function. These data are exciting because there are FDAapproved treatments to prevent autoimmune B-lymphocyte responses. Also, these findings may be broadly applicable to other types of brain injury.

Financial Support: NIH-NINDS (R01 and R21 to MSB), NIH-NINR (K99 to KPD)

It has long been recognized that stroke patients are at high risk for complications due to dysphagia. Performing a swallow screen prior to the administration of any food, fluids or medications has been a standard of care for stroke patients since 2007 and included in the ASA/AHA guidelines (Stroke 2007 (Stroke , 2009 (Stroke , 2013 . "Assessment of swallowing before the patient begins eating, drinking, or receiving oral medications is recommended. A swallow screen should be performed in the first 24 hours after stroke, preferably by the speech language pathologist. Stroke patients should be kept NPO until the screen has been performed (Class I, Level of Evidence B)." In January, 2013, screening for dysphagia for stroke patients became a CMS Core Measure. Although the majority of patients are "stroke-alert" patients who come in through the ED, patients may also be admitted to "non-stroke" services and "non-stroke" nursing units. For these "off-service" patients, the likelihood of staff missing this requirement is very high unless a consistent process is in place.

To address this standard, an interdisciplinary team involving nursing, physicians, speech language pathology (SLP), and information technology built a stroke swallow screening tool in EPIC that the nurse completes for all acute stroke patients. The process includes the swallow screen, "Best Practice Advisories (BPAs)" for consulting SLP and for an NPO order. The swallow screen is scored as "pass" or "fail" and "BPAs" are entered into EPIC according to a "standing order" mode.

The dysphagia screening initiative increased stroke screening compliance from 39.3% to 85.2% (p<0.001).

This quality improvement initiative using a nurse-administered swallow screening tool for acute stroke patients followed by rapid bedside swallow evaluation by a SLP has resulted in improved screening compliance, more timely SLP consults, decreased delays in diet orders for patients who pass their swallow screens and enhanced patient satisfaction.

Therapeutic hypothermia (TH) improves outcomes in various experimental stroke models, especially after ischemiareperfusion injury. We investigated the clinical and radiological effects of TH in patients who had intra-arterial recanalization for acute ischemic stroke.

This is a prospective cohort study at 2 tertiary stroke centers performed over 2 years. We enrolled patients with acute anterior ischemic stroke with an initial NIH Stroke Scale ≥ 10 and a recanalization score of ≥ TICI 2b (thrombolysis in cerebral ischemia). Patients at center A underwent a mild hypothermia protocol which included intubation/mechanical ventilation, 2 days of hypothermia (target temperature: 34.5°C) and 2-day rewarming period. Patients at center B were treated according to guidelines without hypothermia. Cerebral edema, hemorrhagic transformation, good outcome (3month modified Rankin scale ≤ 2), mortality, and safety profiles were compared between centers. Potential variables at baseline and during therapy were analyzed to evaluate independent predictors for good outcome using multivariate regression analysis.

Baseline demographics were well balanced between normothermia (n=36) and hypothermia centers (n=39). The hypothermia group had less cerebral edema (p=0.001) and hemorrhagic transformation (p=0.016), and better outcome (p=0.017). Mortality (p=0.855), need for hemicraniectomy (p=0.629), and serious adverse events (SAE, p=0.089) were not different between groups. After adjustment for potential confounders, TH (OR 3.0, 95% CI 1.0-8.9, p=0.047) and distally involved vessel (OR 7.3, 95% CI 1.3-40.3, p=0.022) were the independent predictors for good outcome. Absence of cerebral edema (OR 5.4, 95% CI 1.6-18.2, P=0.006) and the absence of medical complications (OR 9.3, 95% CI 2.2-39.9, P=0.003) were also independent predictors for good outcome during therapy.

Our results suggest that TH effectively reduces risk of cerebral edema and hemorrhagic transformation after recanalization. In addition, extended period of hypothermia and rewarming, airway protection and aggressive medical treatment may be important for successful use of TH to improve outcomes after stroke.

Hemorrhagic transformation (HT) is one of the complications in acute ischemic stroke and it can lead to poor functional outcome and increase the risk of mortality. Blood-brain barrier permeability (BBBP) imaging has been proposed as imaging predictors for HT. We aimed to investigate whether the sensitivity and specificity of BBBP for prediction of HT are different depending on the type of reperfusion therapy.

Forty six subjects hospitalized with acute cerebral infarction within internal carotid artery or middle cerebral artery who received Intravenous (IV) tPA, mechanical endovascular thrombectomy or both were included in this study. For each subject, baseline demographic characteristics, National Institutes of Health Stroke Scale (NIHSS), thrombolysis in cerebral infarction (TICI) grade before and after reperfusion therapy, BBBP on perfusion CT (PCT) and HT were evaluated.

Initial BBBP above threshold and use of mechanical endovascular thrombectomy were significantly associated with hemorrhagic transformation in multivariate regression analysis. The sensitivity and specificity of initial BBBP measurements as the predictor of HT in overall 46 patients were 80% and 71%, respectively. The sensitivity and specificity of BBBP in patient who received only IV tpa were 75% and 64% whereas 100% and 80% in patient who received only mechanical endovascular therapy. The sensitivity and specificity of BBBP in patient who received IV tPA with or without mechanical endovascular therapy were 77% and 67%, respectively whereas 86% and 76% in patient who received Mechanical endovascular therapy with or without IV tPA.

Prediction of HT after reperfusion therapy with BBBP assessed by PCT is more sensitive and specific in patients who received mechanical endovascular therapy compared with patients who received IV t-PA.

Engorged transmedullary veins (TVs) on Susceptibility weighted imaging are well recognized in patients with acute ischemic stroke, and may represent a state of reduced cerebral perfusion with elevated oxygen extraction fraction. We hypothesized that asymmetric TVs (ATV) in the ischemic hemisphere is associated with poor clinical outcome in patients with acute middle cerebral artery (MCA) infarction.

A consecutive 133 patients with acute MCA infarction within 6 hours from onset were included for the analysis. The numbers of TVs were counted in each hemisphere, and ATV was defined as the difference in number of TVs more than 5.

The lesion growth was defined as greater than 10% increase in infarct volume on follow up DWI. Early neurological deterioration (END) was defined as a fall in ≥ 4 on NIHSS within 48 hours after admission. The degree of collateral flow was dichotomized as good or poor based on the ASITN/SIR Collateral Grading System. 

ATV may be a surrogate marker of poor collaterals, predicting infarct growth and neurologic deterioration.

Neurological sequelae of streptococcus pneumoniae meningitis are well known. Parainfectious neurovascular involvement leading to ischemic brain damage is less frequently reported and there is a paucity of data regarding treatment options. We report the use of intra-arterial (IA) vasodilators in the treatment of meningitis associated vasoconstriction.

Two patients with streptococcal pneumoniae meningitis admitted to the neurologic intensive care unit. Medical records were reviewed; including medications, imaging studies and laboratory data.

Patient one was a 38 year-old man. Despite antibiotics, "triple H" therapy and steroids, neurologic exam worsened and MRI showed new infarcts. Nimodipine and milrinone were also used for presumed vasospasm. Cerebral angiogram showed severe basilar and bilateral internal carotid artery (ICA) vasoconstriction on day nine from symptom onset. Total of 20mg IA verapamil was infused with moderate improvement. Hospital course was significant for increased transcranial doppler (TCD) velocities and worsening exam, prompting a second angiogram on day twelve , which showed moderate vasoconstriction of the basilar and bilateral ICAs and a total of 24mg IA nicardipine was infused. Patient two was a 57

year-old man. After receiving antibiotics, "triple H" therapy, steroids and nimodipione there was radiographic evidence of new infarcts, prompting intrathecal nimodipine. On day 12, cerebral angiogram showed severe narrowing of bilateral ICAs with significant improvement in flow after IA nicardipine. Decline in neurologic exam on day 15 prompted reimaging, which showed new ischemic infarcts. Repeat angiogram and IA nicardipine led to mild improvement in vessel narrowing. After IA vasodilators, both patients' neurologic exam stabilized and there was reduction in TCD velocities.

While medical management of pneumococcal meningitis has been well defined, little is known regarding the management of neurovascular complications. Based on limited data, risk of vasoconstriction occurs two weeks after symptom onset, later than other vascular entities. When there is progressive neurologic deterioration with standard medical care, IA vasodilators may be considered.

Introduction: Although hemicraniectomy (HC) has been found to improve mortality in large hemispheric infarction, hypertonic saline (HS) is often used in the initial management of cerebral edema. The goal of this study is to identify predictors of HC among patients with large Middle Cerebral Artery (MCA) infarction treated with HS.

Methods: Patients admitted to RUMC with ischemic stroke and enrolled into the RUMC stroke database between 1/2011 and 3/2013 were studied. Patient demographics, neuroimaging characteristics, and laboratory results were abstracted through chart review. Univariate analysis was performed to determine the relationship of these factors to hemicraniectomy in patients treated with HS.

Results: A total of 31 patients with MCA stroke were treated with HS. The mean age was 62.8 years and 22 were female. HC was performed in 11 patients. Predictors of HC were fever (p=0.01), intubation (p=0.007), Insulin use ( p=0.02), and internal carotid artery occlusion (p=0.03). Age and glucose >180 were of borderline significance (p=0.06, respectively). Traditional predictors of HC including midline shift >10mm and hemorrhagic transformation were not significant.

Conclusion: Fever, intubation, insulin use and ICA occlusion were found to be significant predictors of HC among patients treated with HS, while traditional predictors such as midline shift and hemorrhagic transformation were not. Future studies should investigate how specific patient and radiographic characteristics influence medical decision-making in the determination of surgical candidacy and their impact on surgical outcomes.

Transcranial doppler ultrasonography (TCD) is a tool commonly employed for the management of subarachnoid hemorrhage related vasospasm in the intensive care unit. Using TCD for monitoring cases of vascular compromise in subdural empyema has not been established.

To utilize transcranial dopplers for early intervention in a case of subdural empyema associated with cerebral vasospasm. Case Summary: 21 year old Hispanic male presented with left hemiparesis and fever for one week. Evaluation revealed right hemispheric subdural empyema and diffuse right frontal cerebritis. Further workup showed complete opacification of maxillary and bilateral paranasal sinuses without sinus wall destruction and right maxillary molar dental caries with abscess. Debridement, and maxillary tooth extraction was done. TCD showed diffuse increased velocities correlating with moderate vasospasm. Patient underwent hemicraniectomy for irrigation and washout of the empyema. Postoperative TCD showed resolution of the vasospasm. When increased left sided weakness was noted, a repeat TCD showed recurrence of mild vasospasm and hyperemia in bilateral middle cerebral arteries. The patient underwent re-exposure and re-accumulation of the subdural empyema was noted

The lumbar puncture revelaed a bacterial infection. MRI and MRA brain with and without contrast showed a right hemispheric subdural empyema and diffuse right frontal cerebritis. CTA head and neck showed diffuse M1 and A1 vasospasm more prominent on the right side. TCD demonstrated diffuse increased velocities correlating to moderate vasospasm. While the immediate post operative TCD showed resolution of the vasospasm, repeat study revealed recurrence of mild vasospasm and hyperemia in bilateral middle cerebral arteries.

Presence of purulent material around cerebral vasculature resulted in increased velocities suggesting vasospasm which was also evident on CT angiography. This resolved with washout and recurred with re-accumulation of empyema. Vascular irritation by purulent material and subsequent spasm and narrowing may be the mechanism whereby strokes occur in patients with subdural empyemas.

We proposed to examine changes in NIH stroke scale (NIHSS) from admission to discharge in a cohort of stroke patients in the Intensive care unit (ICU) and its relation to length of stay and discharge disposition.

We studied 77 acute stroke patients admitted to the ICU in an academic medical center. We extracted the admission and discharge NIHSS for each patient and calculated the difference. We analyzed the overall and ICU lengths of stay (LOS) and discharge disposition. We tested for correlation between the NIHSS change and LOS and discharge disposition using conventional statistical methods (chi square tests, Pearsons correlation co-efficients and regression analysis) and Shewhart control charts.

There was an average improvement in NIHSS of 1.43. Shewhart control charts revealed no special cause test 1 statistical outliers. Ischemic stroke cases showed wide fluctuations in NIHSS compared to intracerebral hemorrhage, suggesting a less well-controlled process. NIHSS difference was found to be related to ICU LOS, individually (p=0.036) and independent of the other stroke risk factors (p=0.014). Changes in NIHSS also appeared to be linked to the discharge disposition with 40 out of the 45 patients with improvement in NIHSS being discharged home or to rehabilitation compared to 11 out of 23 patients with worsening NIHSS and 4 out of 9 patients with no change in NIHSS (p = 0.0012).

Conclusions NIHSS change appears to be related to the ICU LOS and discharge disposition. Application of Shewhart control charts may provide valuable additional information providing us with opportunities to improve health care delivery. In this study, control charts found wider variation in NIHSS over hospital stays for ischemic stroke compared to intracerebral hemorrhage, suggesting that for ischemic patients there are more opportunities to secure a good outcome as well as more risks of clinical worsening, depending on care provision.

Adenosine (Ado) is a ubiquitous, endogenous purine nucleoside with pleiotropic effects, including modulation of neuronal activity and neuroprotection. These effects are due to post-synaptic hyperpolarisation and pre-synaptic inhibition of neurotransmitter release. Low levels of adenosine could lead to brain hyperactivity. Hypoxic astrocytes are capable of rapidly producing high concentrations of extracellular Ado. Consequently, astrocytes may be the source of adenosine in the brain. We sought to determine if high oxygen levels (95% O 2 , 5% CO 2 ) depress adenosine levels and could account for symptoms associated with hyperoxia.

Primary, mixed cultures were established from the cerebral hemispheres of 1-3 day old male Sprague dawley rat pups.

After 5-7 days in vitro (DIV), cultures were purified based on differential adhesion and a high purity (>95%) ascertained with GFAP/DAPI immunolabelling. Astrocytes were trypsinized after 11 DIV, attached to microcarrier beads, and kept in spinner flasks. After 23 DIV, O 2 tension in the culture was raised to 720 mmHg (95% O 2 ) by flushing flasks with compressed oxygen. We measured pO 2 in our culture continuously utilizing a fluorometric technique. Steady-state samples of supernatant were acquired after 20 min each at normoxic and hyperoxic levels. Samples were cleaned, the eluent nitrogen evaporated, reconstituted and analyzed with HPLC.

During high (n=6) levels of oxygen (95%O 2 , 5%CO 2 ), the concentrations of adenosine, inosine and hypoxanthine were significantly (p<0.05) suppressed compared to normoxic (n=6) levels of oxygen(20%O 2 ).

The attenuated levels of adenosine seen during high levels of oxygen could lead to neuronal disinhibition and may explain the clinical symptoms of hyperoxia, i.e. seizures, altered mental state, loss of consciousness. In addition, earlier investigations using 95% O 2 in hippocampal and cortical slices may need re-evaluation. This level of O 2 , based on our preliminary observations, would severely suppress resting Ado levels, thus affecting, in a heretofore unrecognized fashion, both neurophysiologic and arteriolar responses.

Acute stroke has become the leading cause of disability in the USA and carries a major socio-economic burden. Based on the Nationwide Inpatient sample, the largest publicly available all-payer inpatient database in the United States, we sought to establish the national estimates for the frequency of complications, the need for additional procedures and the discharge dispositions for patients with acute ischemic stroke treated with either intravenous tPA or intra-arterial therapy (IA).

Discharge records from the Nationwide Inpatient Sample between 2006 and 2009 were obtained and admissions for acute ischemic stroke identified using the ICD-9 discharge diagnosis code. Patients undergoing IA therapy were identified using the ICD-9 procedure code 39.74, while the procedure code 99.10 or diagnosis code V45.88 were chosen to identify patients receiving tPA. Data were analyzed with Microsoft Excel and SPSS.

In patients who had undergone IA therapy, the rate of intracranial hemorrhage was significantly higher than in patients treated with IV tPA only. The need for additional procedures (craniotomy, mechanical ventilation and tracheostomy placement), was likewise significantly higher. The rates of PEG tube placement, aspiration pneumonia or deep venous thrombosis did not differ significantly between both treatment groups. The rate of routine discharges after IV tPA were significantly higher compared to IA therapy, while the rates of discharge to short-term care, SNF rehab and hospice were comparable. The rate of fatal demise was significantly higher with IA therapy.

Based on this retrospective study IA therapy has a higher morbidity and mortality than IV tPA. However, the sample database does not account for stroke severity. Thus, a bias to more severe strokes in the IA group can't be excluded. Knowledge of this data will aid physicians in leading discussions with patients and families about treatment decisions in the setting of an acute stroke.

Acute ischemic stroke (AIS) due to the occlusion of internal carotid artery terminus (ICA-T) is associated with no or poor recanalization in both intravenous (IV) and inra-arterial (IA) thrombolysis and poor outcome. Objectiveis isto evaluate the clinical and radiographic outcome of emergent thrombectomy of the ICA-T occlusion in patients with AIS.

Consecutive patients with ICA-T occlusion with AIS were enrolled. Patients demographic including time of symptoms onset, presenting stroke scale and peri-procedural information were collected. Additionally, a 90 days outcome was measure using National Institute of Health Stroke Scale (NIHSS) and modified Rankin Scale (mRS).

7 patients with median age of 53 median NIHSS 15 was diagnosed with AIS and ICA-T occlusion. Of 7 patients; one was failed and 6 were not an intravenous thrombolytic candidate. All patients received small dose of IA TPA except two followed by thrombectomy using Solitaire device resulted in complete recanalization. The median time from groin puncture to middle cerebral artery microcatheter placement was 29.3 minutes and recanalization was 61.9 minutes. Immediate more than10 point reduction in NIHSS) was observed in 5 cases including two with NIHSS 0. 90 days good outcome was observed in 6 patients (mRS 0 in 2, mRS 1 in 2, mRS 2 in 2). Poor outcome (mRS 4) was observed in one 28 years women who presented with NIHSS 18 with unknown time of onset, developed malignant edema required hemocraniactomy, ventriculostomy catheter and pentobarbital com. However this patient has now begun to walk with walker, has fluent speech and intact cognition.

Primary thrombectomy with Solitaire device for ICA-T occlusion in acute ischemic stroke patients who are either failed or not an IV thrombolytic candidate not only achieve complete recanalization but also associated with good clinical outcome. Therefore, primary thrombectomy should be offered to all eligible acute ischemic stroke patients.

Malignant cerebral infarction has high mortality. Hypertonic 3% saline (3%HTS) is commonly used to control cerebral edema, despite lack of evidence regarding efficacy. The purpose of this study is to determine if 3%HTS infusion reduces likelihood of brain herniation in malignant cerebral infarction.

IRB approval was obtained. Cases of malignant cerebral infarction (NIHSS>15, infarct volume >80cc) were identified between 2006-2012, that were either treated with 3%HS ("treatment") or no edema-modifying therapy ("controls"). Data collection included gender, age, medical history, infarct volume, vitals, serum sodium levels before and during therapy, dose and duration of therapy, and outcomes. The outcomes were classified as "good" in cases of survival without hemicraniectomy, or "bad" in cases of hemicraniectomy or in-hospital death. We applied Chi-square and logistical regression analyses.

Total 260 cases were reviewed: 54 treatment and 206 controls. Among controls,133/206 (51%) had good outcome, and 73/206 (28%) had bad outcome. Among treatment cases, 24/54 (9%) had good outcome, and 30/54 (12%) bad outcome. There was a significant (p=0.007) association between treatment and bad outcome. When large volume strokes (>140cc) were separately analyzed, there was no significant difference (p=0.39) in outcomes between treatment and controls. When logistical regression was applied, and the effects of stroke volume and other contributing factors was accounted for, there was a trend (OR 1.914, CI [0.99-3.71] p=0.0545) towards bad outcome with treatment. Baseline serum sodium, change in serum sodium, peak serum sodium, dose and duration of treatment were not statistically correlated with outcome.

This study suggests that 3%HS is not effective, or may be harmful, in the management of cerebral edema in malignant cerebral infarction. This study may be contaminated by biases of retrospective analysis. However, this data adds to the little that is known about the efficacy of 3%HS in the management of cerebral edema.

Malignant middle cerebral artery (MCA) infarction occurs in 10% of all ischemic strokes and is associated with high mortality rates. Early decompressive craniectomy reduces mortality rates and improves functional outcomes.

The purpose of the study is to evaluate the efficacy of decompressive craniectomy in patients with malignant MCA syndrome due to MCA or internal carotid artery (ICA) territory infarction. This is a prospective, observational study from 2000-2012 of 151 patients (92 men and 59 women (mean age 54.76 years, range 34-83 years) treated by decompressive craniectomy for MCA/ICA territory infarction at a comprehensive stroke center. 158 underwent conservative care without surgery. Neurological/ radiological findings, modified Rankin Stroke Scales (mRSS) and National Institutes of Health Stroke Scale (NIHSS) at presentation, before and after surgery were recorded. Clinical outcomes were assessed using the Glasgow Outcome Scale (GOS). Mortality rates were assessed. A mRS score of > 3 was considered to represent a poor outcome.

Mortality rates were significantly lower in the craniectomy group as compared to conservative group: one month and 3 month mortality rate being 38.7% and 57.6 % respectively, as compared to 52.4% and 78.6 %, respectively, in the conservative care group (p <0.001). Age did not statistically differ between the two groups. Elderly patients beyond the age of 70 years tolerated the procedure well. Average preoperative NIHSS was 24.4±8.24. Functional independence (mRS <2) was achieved in 18.4 % at 6 months.

Decompressive craniectomy is effective for patients with a malignant MCA infarction regardless of their age. The rate of mortality is relatively high and most survivors experience severe disabilities with only 18.4 % achieving functional independence. Careful observation of prognostic factors other than age should be considered and treatment should be individualized in elderly patients with malignant middle cerebral artery infarctions.

Middle Cerebral Artery (MCA) territory strokes may leave patients unable to swallow safely. Decisions regarding artificial nutrition and goals of care often arise in patients with severe strokes leading to dysphagia. This study determined some predictors of early transition to palliative level of care among patients with acute ischemic MCA stroke with dysphagia.

This is a retrospective cohort study. Demographic and clinical data of patients presenting to Hartford Hospital with an acute ischemic stroke between January 2005-December 2010 were gathered utilizing the Stroke Center at Hartford Hospital Database. The 236 patients included were divided into "early transition" and "not transitioned" to palliative care cohorts. Primary outcome was transition to palliative care. Significant factors associated with an early transition to palliative level of care in univariate analysis were then entered into a multivariate logistic regression analysis to identify potential independent predictors of early transition to palliative level of care.

79 patients (34%) were transitioned to palliative level of care after failing the first swallow evaluation . Factors predictive of an early transition to palliative level of care included advancing age (p <0.001; OR: 1.10) , left MCA infarct (p = 0.039; OR: 0.417), a high admission NIHSS score (p = 0.017; OR: 3.038), administration of intra-arterial tPA (p <0.001; OR: 7.106) and the inability to be assessed on the 1 st swallow evaluation (p <0.001; OR 0.053).

Independent predictors of an early transition to palliative level of care among patients with an acute MCA territory stroke and dysphagia included advancing age, a left MCA infarct, a high NIHSS score on admission, administration of intraarterial tPA and the inability to be assessed on the 1 st swallow evaluation. This information may guide discussions with families of patients with MCA territory strokes regarding artificial nutrition and goals of care.

Staff. The funding body did not influence the design, collection, analysis and interpretation of data as well as the decision to submit the manuscript for publication.

Stroke is associated with a high risk for death, especially in the first few weeks after the attack. However, few studies have been published on the specific causes of death after an acute ischemic stroke. We assessed the hypothesis that a high percentage of the deaths occurring during an acute stroke admission are direct consequences of the stroke (herniation or hemorrhagic transformation) or related to care withdrawal.

We performed a post-hoc analysis of prospectively collected data of consecutive patients admitted to a Brazilian tertiary hospital with acute ischemic stroke from February 2009 to April 2012. Death causes were extracted from patients' charts and death reports and discussed with the attending team.

We evaluated 424 patients with acute ischemic stroke ( 

In conclusion, although, direct stroke consequences were frequent causes of death in our series, infections were still common and preventable causes of death in patients with ischemic stroke. Care withdrawal should not be ignored as a cause of death in patients with stroke, being a potential important source of bias in observational stroke studies. 

Qualitative research using phenomenological inquiry. This is a type of qualitative research wherein a group of people's experiences of a certain phenomenon (stroke support group) provides a platform for the methodology of this study. Semistructured focus groups were conducted with 15 members of a stroke survivor support group. Interview notes, surveys, reflections, follow-up from support group members, and transcription from the interview were reviewed and coded by the authors. The key phrases were grouped to identify themes associated with the needs of stroke survivors. Additionally, 14 members of the stroke group completed a survey where they conveyed their opinions about recurrent stroke.

15 people (6 stroke survivors, 5 care-partners, and 4 professionals) participated in interviews. Themes that emerged included: treatment and self-identity. Within the theme of treatment the following sub-themes developed: a) treatment from health care professionals, b) treatment from family members, and c) a plan for treatment after leaving the hospital. The theme of self-identity did not contain sub-themes

Understanding the patient perspective is critical for improving patient centered outcomes. There is a need to address themes on optimizing treatment during the early recovery phase of stroke. This research has the potential to improve short-term and long-term patient outcomes through enhancing patient care, and buy-in of stroke recovery treatment by both patients and families.

Financial Support: None S144 Neurocrit Care (2013) 19:S1-S331 

Bacterial CNS infections have a significant prevalence and a high risk for severe, life-long disability or mortality. High-dose intravenous antibiotics remain the first treatment strategy. However, patients with complicated, persistent or recurrent infections, especially device-related, receive intraventricular or intrathecal (IVT/IT) antibiotics. We analyzed the English literature to describe IVT/IT treatment regimens and their pharmacodynamic treatment responses.

Using PubMed, Embase, Ovid-MEDLINE, internet-JAMA, Web-of-Science, online resources/textbooks, and Google-Scholar the literature was searched from 1943-2013 for IVT/IT antibiotic treatment with reported CSF drug levels. 1533 publications were retrieved; after applying selection criteria 50 studies with 141 patients were analyzed. We collected clinical/microbiological/treatment/CSF antibiotic concentrations characteristics and microbiological/patient outcome.

We identified 3 prospective-randomized-trials (38patients; 1 catheter-induced ventriculitis, 2 meningitis) and 48 case reports/series. Diagnoses were ventriculitis (n=69; 49%), meningitis (n=62; 44%), ventriculomeningitis (n=10; 7%). Cause of infection was a) not delineated (n=51; 36%); b) CNS device (n=63; 70%); c) exposure of intracranial content (n=15; 17%), d) immunosuppressive state (n=3;3%). 115/141 (82%) patients received IVT, 23 (16%) IT. A pathogen was identified in 94% (n=132) with 20 different organisms, most commonly Staphylococcus species (38%), Enterococcus faecalis (10%), and Pseudomonas aeruginosa (8%). 13 different antibiotics were administered, most commonly gentamicin (43/141; 31%; 2.5mg/24hrs), vancomycin (23; 16%; 10mg/24hrs), colistin (15; 11%; 2.4mg-5.22mg/12-24hrs), teicoplanin (not US approved; 12; 8.5%; 10mg/24hrs), and amikacin (11; 8%; 5mg/24hrs).270 CSF drug levels were reported with 206 (76%) trough levels and 64 (24%) peak levels. The CSF did not sterilize in 8/145 (6%); no information was available in 40 (28%).

Common treatment pattern of local CNS treatment approaches were identified. Meningitis and ventriculitis are the leading indications; intraventricular the preferred route; Staphylococcus, Enterococcus and Pseudomonas the most common pathogens; and CFS drug levels varied among different dosing regimens of gentamicin, vancomycin, colistin, and amikacin.

Although subdural hematoma (SDH) is common in neurocritical practice, little is known about SDH patients requiring prolonged mechanical ventilation (PMV). Therefore, we aimed to determine predictors of PMV and relationship to outcome in a large cohort of patients with SDH.

Adult patients admitted to Rush University neurointensive care unit with SDH between 1/2009 and 3/2012 were retrospectively reviewed. Duration of intubation and pulmonary complications (pneumonia, reintubation, tracheostomy placement) were reviewed. Demographics, comorbidities, treatment, discharge disposition and length of stay (LOS) were identified. PMV was defined as duration of intubation > 72 hours. Univariable and multivariable analyses were performed to identify predictors of PMV and association with outcome amongst survivors with SDH.

Of 288 

Among patients with SDH, alcohol abuse, admission GCS <15, and surgical treatment are predictive of PMV. PMV is associated with pulmonary complications, increased hospital LOS, and unfavorable discharge destination. Future studies should investigate the role of early tracheostomy in high risk patients and impact on outcomes.

The apnea test is a crucial component of the clinical diagnosis of brain death. Significant hypoxemia, hypotension and cardiac arrhythmias may occur, sometimes requiring premature termination of the test. The purpose of this study was to perform a contemporary re-evaluation of the safety of the apnea test with the current advancements in critical care.

We performed a detailed chart review of consecutive brain dead patients who underwent an apnea test from 2008 to 2012.

Out of 63 patients, 33 were men (52.4%). Mean age was 46.4 years. In all but 4 patients (93.7%), the apnea test was performed by a neurologist. Infiltrates on chest radiographs were present in 34 (54%). Seven patients (11.1 %) had chest tubes, 6 of which were associated with polytrauma. Echocardiograms were obtained in 47 patients (74.6%) and 18 patients (38.3%) had regional wall motion abnormalities (IQR 41-65%). Fifty patients (79.4%) were on vasopressors prior to apnea test. Median FiO2 was 0.5 (IQR 0.4-0.6) and PEEP was 5 (IQR 5-10). After apnea test median pO2 was 306 mmHg (IQR 121-389). Apnea test had to be aborted in only 1 patient with aSAH, due to desaturation (FiO2 0.9-1 with 10 cmH2O of PEEP before ventilator disconnection). Mild hypoxemia occurred in 3 others without any consequences. Mild hypotension occurred in 9 patients (14.3%) and was easily managed by an increase in the vasopressor infusion. There were no instances of major cardiac arrhythmias.

We found a major decrease in the number of aborted or not attempted apnea tests compared to previous studies. Apnea test using the oxygen-diffusion technique is safe, including patients with polytrauma, provided the pre-requisites are met.

Mannitol in has been shown to cause cellular stress inducing apoptosis in bovine endothelial cells in a dose-dependent manner as well as activate inflammatory and coagulation pathways leading to intravascular thrombosis. Some researchers have demonstrated that mannitol may cause hemagglutination or dehydration causing venous thromboembolism (VTE) though the data supporting these phenomena are scarce. Though venous thrombosis or phlebitis extending from the injection site is listed as a possible side effect of mannitol, there is a paucity of data on whether clinical VTE in hospitalized patients can be attributable to mannitol. Interestingly, we previously demonstrated an increased risk of PICC-venous thrombosis in neurological intensive care unit patients who received mannitol, but not hypertonic saline. In this study we investigate the association of VTE in a high risk group of brain injured patients requiring osmotic therapy.

Consecutive patient admitted from January 1 st 2005 through December 31 st 2012 requiring osmotic therapy were included. Cumulative mannitol dose, confounders and thrombosis (any large or deep vein thrombosis or pulmonary embolus) were documented. Descriptive statistics were used to quantitatively evaluate the cohort (Tables 1 and 2) . Propensity score matching, predicting the probability of exposure to greater than the median mannitol dose, was used to reduce selection bias and control for confounding. Risk of thrombosis was evaluated with a Cochran-Mantel-Haenszel OR and multivariable conditional logistic regression.

The proportion of patients diagnosed with a thrombosis did not significantly change over time; however the use of mannitol declined while the proportion of patients given hypertonic saline increased (Figure 1 ). The odds of thrombosis was not different among those receiving mannitol compared to those that received only hypertonic saline (1.11 (95% CI 0.59-2.1)) even after adjusting for year of treatment (1.2 (0.71-2.0)) [ Table 3 ].

We found no evidence that mannitol use was associated with increased risk of VTE compared to 3% hypertonic saline.

The use of pharmacologic venous thromboembolism (VTE) prophylaxis in patients with acute intracranial hemorrhage has been controversial due to concerns of hemorrhage expansion. We compared mechanical to pharmacologic VTE prophylaxis.

In a prospective study of intracranial hemorrhage patients (subarachnoid hemorrhage N=117; intracerebral hemorrhage N=128 and subdural hemorrhage N=133) conducted between 7/2008-11/2011, we compared bleeding complications, VTE rates and 3-month functional outcomes between patients who received pharmacologic VTE prophylaxis (either heparin or enoxaparin) plus compression boots versus mechanical prophylaxis (compression boots) alone.

Of 378 patients, 165 (70%) received pharmacologic prophylaxis (heparin BID N=22 [6%]; heparin TID N= 107 [28%]; enoxaparin N=136 [36%]) in addition to compression boots, while 95 (25%) received compression boots alone. The median time to initiation of pharmacological prophylaxis was the same as mechanical prophylaxis alone (4 days post hemorrhage onset; 2 days from surgical intervention). Compared to mechanical prophylaxis alone, those who received pharmacologic prophylaxis were younger (median age 60 versus 65, P=0.002), more likely to have a SAH (41% versus 18%, P<0.0001), less likely to have SDH (25% versus 73%, P<0.0001) and less likely to be DNR/comfort care (14% versus 18%, P<0.0001). Admission GCS and APACHE 2 scores did not vary between groups. There were no significant differences in bleeding events (EVD associated hemorrhage, ICH expansion, new ICH, new SDH, SDH reaccumulation, bleeding at the craniotomy site, GI bleeding or anemia requiring transfusion) or thombotic events (DVT or PE) between those receiving pharmacologic prophylaxis versus mechanical prophylaxis alone. Pharmacologic prophylaxis was significantly protective against death at 3 months after adjusting for age, admission GCS, bleed type, and DNR/comfort care status (aOR 0.2, 95% CI 0.1-0.6, P=0.002).

In intracranial hemorrhage patients, pharmacological VTE prophylaxis is not associated with higher bleeding risks and predicts improved 3-month mortality rates compared to mechanical VTE prophylaxis alone.

Both unfractionated heparin and enoxaparin are commonly used for venous thromboembolism (VTE) prophylaxis in patients with intracranial hemorrhage, yet the risks and benefits of each are not well delineated. We hypothesized that enoxaparin or heparin TID are superior to heparin BID in improving outcomes without increased risk of bleeding complications.

In a prospective study of 265 intracranial hemorrhage patients (subarachnoid hemorrhage [SAH] N=108; intracerebral hemorrhage [ICH] N=91 and subdural hemorrhage [SDH] N=66) conducted between 7/2008-11/2011, we compared bleeding complications, VTE rates and 3-month functional outcomes between patients who received heparin 5000 u SQ BID (N=22), heparin 5000 u SQ TID (N=107) and enoxaparin 40 u QD (N=136).

Compared to those who received BID or TID heparin prophylaxis, patients who received enoxaparin were younger, less likely to have renal insufficiency, more likely to have SAH, and less likely to have SDH ( all P<0.05). Enoxaparin was initiated sooner after intracranial hemorrhage onset than heparin dosed BID or TID (median 4 versus 6 and 4.5 days, respectively, P=0.011). Admission GCS and APACHE-2 scores did not vary between groups. Patients receiving heparin BID were more likely to have GI bleeding (5% versus heparin TID 0% and enoxaparin 2%; P=0.007) and ICH expansion (9% versus 2% and 1%, P=0.022), though other bleeding rates (EVD associated hemorrhage, aneurysm rebleeding, new ICH or SDH, SDH reaccumulation, anemia requiring transfusion) and VTE rates did not vary between groups. At 3months, compared to heparin TID and enoxaparin, heparin BID patients had worse Barthel Index independent activities of daily living (P=0.027) and cognitive outcome (Telephone Interview of Cognitive Status, P=0.011).

In intracranial hemorrhage patients, compared to heparin 5000 u BID, enoxaparin and heparin 5000 u TID were not associated with any increased bleeding risk, and were associated with better 3-month functional and cognitive outcomes.

The epidemiology of circulatory shock after status epilepticus (SE) is unknown. We sought to determine its incidence, risk factors, and effect on case-fatality. We tested the hypothesis that shock was not associated with higher case-fatality after admission to the Intensive Care Unit (ICU).

Retrospective multi-center cohort study of 83 ICUs in the United States from 2003-2008. SE patients >17 years of age admitted to an ICU. Shock was defined as systolic blood pressure (SBP) <90 mm Hg or mean arterial pressure (MAP) <70 mm Hg, or vasopressor requirement to keep SBP>90 mm Hg or MAP>70 mm Hg, and duration >1 hour during admission or ICU-stay. We ascertained cases of shock after admission and during ICU-stay.

Total of 770 patients. Mean age was 53(SD 16), 434(56%) were female, 499(67%) were white, 463(61%) were functionally independent, and median APACHE-II was 15 (interquartile-range[IQR] 10-21). Overall case-fatality for SE was 16%(125/770). The incidence of shock upon ICU admission was 15%(115/770) and 16%(126/770) during ICU stay. Of those with admission shock, 58%(67/115) improved and 42%(48/115) did not (p<.0001). SE-shock patients in whom shock resolved, had a case-fatality of 22%(15/65) vs. 28%(17/60) when it did not resolve(p<.0001). In a multivariable analysis controlling for disease severity, ICU organ dysfunctions, and hospital characteristics, the following were independent predictors of case-fatality: age, admission non-independent status, APACHE-II >15, hematological dysfunction, and shock during ICU stay (aOR 4.3,95%CI 2.3-7.9,p<.0001) but not admission shock (aOR 1.2,95%CI 0.6-2.3,p=0.6).

Circulatory shock is frequent in SE patients admitted to the ICU. The impact, risk factors, and etiology of shock in SE patients may differ from admission and during ICU-stay. Admission shock may reflect effects of SE specific therapies, whereas shock during ICU-stay may reflect complications of such therapies or critical-illness. Our findings require further study.

Acute lung injury is a frequent and deadly complication in acute brain injury. We sought to identify the incidence and risk factors for ARDS based on the new Berlin Criteria for ARDS in subarachnoid (SAH) and intracerebral hemorrhage (ICH) patients on mechanical ventilation for at least 72 hrs. We hypothesize that ICH and SAH patients who develop ARDS will have worse outcomes at hospital discharge.

This study is a single center retrospective review from 2009-2012, to include ICH and SAH mechanically intubated for at least 72 hrs and within 24 hrs of admission. Demographics, relevant past medical history, severity of illness scales, length of mechanical ventilation, ICU and hospital length of stay were collected. Ventilator settings and measurements were collected on days 1, 3 and 7 days of post intubation, daily chest xrays and arterial blood gas (ABG) results were reviewed for the first 7 days of ICU admission.

To date, our analysis includes 115 patients (ICH=85 and SAH=30) meeting enrollment criteria, of whom 20 (17%) 

In our initial univariate analysis, the incidence of ARDS is lower than expected, and likely decline further with review of echocardiograms. In addition, while less acutely ill and younger, ARDS patients had a higher mortality Additional analysis will include risk factors for ARDS, as well as multivariate analysis for mortality with ARDS.

Hyponatremia is a significant cause of morbidity and mortality in neurologically injured patients. The incidence and severity with relationship to outcomes in this population has not been well described.

This retrospective, multicenter, observational study identified 400 ICU patients, from 17 centers, admitted for ≥48 hours with SAH, TBI, IPH, or intracranial tumors between January 1,2011 and July 31,2012. Data collection included demographics, APACHE, GCS, serum sodium, fluid rate and tonicity, sodium altering therapies, LOS, and mRS upon discharge. Data was collected for the first 21 days of ICU admission or ICU discharge, whichever came first. Hyponatremia severity was defined as mild (Na+ 131-135mEq/L), moderate (125-130mEq/L), or severe (≤124mEq/L).

54% ( 

Hyponatremia is common in patients with acute neurologic injury and infrequently severe as treatment is often initiated early or even before hyponatremia develops. Hyponatremia is further associated increased LOS. Further evaluation is needed to determine if treatment improves outcome.

Little data exists regarding the practice of sodium management in acute neurologically injured patients. This study describes the practice variations, thresholds for treatment, and effectiveness of treatment in this population.

This retrospective, multicenter, observational study identified 400 ICU patients, from 17 centers, admitted for ≥48 hours with SAH, TBI, IPH, or intracranial tumors between January 1,2011 and July 31,2012. Data collection included demographics, APACHE, GCS, serum sodium, fluid rate and tonicity, sodium altering therapies, ICU and hospital LOS, mRS upon discharge. Data was collected for the first 21 days of ICU admission or ICU discharge, whichever came first. Sodium trigger for treatment was defined as the sodium value prior to treatment and response an increase of ≥4mEq/L at 24 hours.

Sodium altering therapy was initiated in 34% (137/400) of patients with 23% (32/137) having Na+ >135mEq/L at time of treatment initiation. The most common indications for treatment were declining serum Na+ (68/116,59%) and cerebral edema with mental status changes (21/116,18%). Median Na+ treatment trigger was 133mEq/L (IQR 129,139) with no difference between diagnoses. Incidence and treatment of hyponatremia was more common in SAH and TBI [SAH (49/106,46%), TBI (39/97,40%), ICH (27/102,26%), tumor (22/95,23%); p=0.001]. The most common initial treatment was hypertonic saline (85/137,62%), followed by oral NaCl (42/137,31%), and fluid restriction (15/137,11%). Among treated patients, 60% had a response at 24 hours. Treated patients had lower admission GCS (12 vs. 14,p=0.02), higher APACHE II scores (12 vs. 10,p=0.001), longer ICU (9 vs. 5,p<0.0001) and hospital LOS (13 vs. 9,p=0.0001).

Sodium altering therapy is commonly employed among neurologically injured patients. Hypertonic saline infusions were the most common first line strategies employed with the majority having a positive response at 24 hours. Further studies are needed to evaluate the impact of various treatments on patient outcomes.

The development of new onset acute kidney injury (AKI) and its impact on outcomes in the subarachnoid hemorrhage population has not been well described. Specific aim: To study the incidence of developing AKI (based on Acute Kidney Injury Network criteria) and its impact on outcomes in the SAH population admitted to a large academic neurocritical care unit.

This is an IRB-approved, retrospective study of patients admitted to a tertiary academic neurologic intensive care unit between 2006-12. The patient cohort included adult (age>18Y) patients admitted following SAH (aneurysmal and non aneurysmal etiologies). Comparisons between the AKI and control group were conducted using the Wilcoxon rank sum test for continuous variables, and the Pearson Chi square test for categorical variables. Confidence intervals for AKI incidence and the odds ratio of mortality for the AKI versus control group were calculated using a bootstrap method. The primary endpoint was the development of AKI since admission to the ICU up to 14 days post bleed. The secondary end point was hospital mortality.

The final cohort of patients included 1253 adults admitted to the neurosciences ICU. The number of patients who developed AKI was 133 (11%). Patients in the AKI cohort were older (median age was 60 Y). Diabetes mellitus was significantly more prevalent in the AKI group (29% Vs 16 % with p<0.001). The AKI cohort had an increased length of ICU stay 14.4 ± 9.3 days Vs 11.2 ± 8.5 days (p<0.001). Patients who developed AKI had a 2.36 (95% CI: 1.51,3.69) increased odds of death compared to the patients who did not develop AKI.

The development of new onset AKI is associated with a significant increased length of stay and significant increased risk of hospital mortality.

Current literature suggests that the routine use of unfractionated heparin (UFH) in the setting of acute brain injury could pose greater risk than benefit. However, acutely brain injured patients with concomitant thrombosis may require anticoagulation. Minimal evidence is available to guide anticoagulation in these high risk patients. This study aims to assess the safety and efficacy of bolus-driven "high dose" and bolus-free "low dose" brain injury heparin drip protocols at a single, tertiary, academic hospital.

This was a retrospective, descriptive, observational study. Patients were identified for study through a computerized order entry review for both heparin drip protocols initiated under the neurology or neurosurgery services between August 2012 and April 2013. Electronic charts were accessed and retrospectively evaluated. The "high-dose" and "low-dose" protocols only differed in the use of a 30 unit/kg bolus dose in the "high dose" protocol. Patients without acute brain injury were excluded from analysis.

Seventy-five acutely brain-injured patients were treated with the high dose protocol (n=31) or the low dose protocol (n=44). The high dose protocol was associated with no new or recurrent thromboembolic events as compared with 4.5% receiving the low dose protocol. Hemorrhagic events were experienced in 16.1% of patients receiving the high dose protocol as compared with 2.3% of patient receiving the low dose protocol. All hemorrhagic events occurred greater than 24 hours after the initiation of the protocol.

The high dose protocol was associated with a smaller percentage of recurrent thromboembolic events. However, the high dose protocol was associated with a larger percentage of hemorrhagic events occurring greater than 24 hours after protocol initiation. Further data collection and analysis are required to determine if certain subsets of acutely brain injured patients would benefit from a bolus-driven UFH protocol.

Valproic acid (VPA) use is common in neuroICU patients. While typically well-tolerated, VPA can cause hyperammonemia. We hypothesize that VPA-associated hyperammonia is common in the neuroICU population and factors such as mechanical ventilation are associated with higher likelihood of developing hyperammonemia.

From a prospective database we identified a cohort of 78 neuroICU patients from 2011-2013 who had received VPA, collected patient characteristics including baseline VPA use, mechanical ventilation and duration, presence of status epilepticus (SE), average daily VPA dose and peak serum ammonia levels. Hyperammonia was defined as serum ammonia >60 μmol/L. Continuous data were compared using student's t-test or Wilcoxon rank sum test depending on data normality, and categorical data were analyzed using chi-squared test. A logistic regression model was used to identifypredictors of hyperammonia associated with VPA use.

Of the 78 patients evaluated, 37 had ammonia levels measured. The overall prevalence of hyperammonemia in VPAtreated neuroICU patients was 21% (16/78). Of those with ammonia checked, 43% (16/37) had hyperammonemia. Mechanical ventilation (81% vs 39%), duration of mechanical ventilation (9.75d vs 2.18d), higher average daily VPA dose (2606mg vs 1362mg) and presence of SE (87.5% vs 38.7%) were more common in the hyperammonia group on univariate analysis. Mechanical ventilation (p=0.037, OR 5.65) and baseline VPA use (p=0.033, OR 4.4) were independently associated with VPA-associated hyperammonemia while SE showed trend towards association (p=0.06, OR 5.19) in multivariate analysis.

VPA-associated hyperammonemia is prevalent in the neuroICU population. Mechanical ventilation and baseline VPA use are independent risk factors associated with developing VPA-induced hyperammonemia . Routine screening for hyperammonemia in neuroICU patients treated with VPA is prudent given the high prevalence of this condition. A larger study is necessary to identify other potential risk factors for developing VPA-associated hyperammonemia in the neuroICU population.

Financial Support: None

The purpose of this study was to describe the use of subcutaneous Heparin 7,500 units every 8h for venous thromboembolism (VTE) prophylaxis in patients who weighted ≥100 kg admitted to a neuroscience units.

This is a retrospective cohort study. In 2010, a protocol was employed for VTE prophylaxis in patients weighing ≥100kg. Dosing changed, from subcutaneous Heparin 5000 units every 8h to Heparin 7500 units every 8h, to account for bodyweight and to lower incidence of VTE. Data collection included baseline demographics, diagnosis, clinically significant bleeding as documented hemoglobin drop ≥ 2 g/dl in 24h or from baseline, pRBC transfusion, an increase in PTT, documented evidence of worsening intracranial bleeding associated with the use of Heparin, VTE, and hospital length of stay. The primary endpoint was incidence of significant bleeding. The secondary endpoint was incidence of VTE.

From July 2010 to November 2013, 124 patients met the inclusion criteria for Heparin 7500 units every 8h (n=88) and Heparin 5000 units every 8h (n=36). The two groups had similar rates of bleeding complications. The incidence of drop in hemoglobin by 2 points within 24 hours was (21% vs. 17%; P=0.63), 2 points from baseline (60% vs. 53%; P= 0.45), pRBC transfusion (23 % vs. 26%); P=0.77) increase of PTT from baseline by 2 times (1.2% vs. 2.8%, P= 0.53). There were no differences in the occurrence of HIT, hematuria, epistaxis, or presence of positive blood NG aspirate and guaiac positive stool. Discontinuation of heparin therapy for association with progressive bleeding was not documented in any of the study patients. No difference in the incidence of VTE was observed (8% vs. 8.3%; P=0.9).

Heparin 7500 units every 8h for VTE prophylaxis in ≥100 kg patients may be safe, but a decrease in incidence of VTE was not demonstrated.

Critically ill patients with neurological pathology can experience elevations in intracranial pressure (ICP) which may be associated with the underlying condition or may occur in response to routine nursing interventions. Nurses who care for critically ill patients with increased ICP must balance the need for oral care to promote oral health with the need to minimize elevations in ICP. The purpose of this study was to describe the relationship between routine oral care interventions and the changes in ICP.

In this non-experimental observational study of the effect of oral care on ICP, 23 patients with a clinical condition requiring ICP monitoring were enrolled over a 12 month period. A total of 33 episodes of oral care provided by neuroscience intensive care nurses were observed and videotaped. Characteristics of the intervention were documented including products used, patient positioning, and duration of the intervention. A 1-5 subjective scale was used to score intensity of oral care. Wrist actigraphy data were collected from the nurses to provide an objective measure of intensity. ICP, cerebral perfusion pressure (CPP), heart rate, respiratory rate, pulse oximetry, systolic, diastolic and mean blood pressures were collected at 12 second epochs 5 minutes before, during and 5 minutes after oral care.

The mixed effect repeated measures ANOVA model indicated a statistically significant increase in ICP in response to oral care (p=0.0031) but no significant difference in CPP. There was no clinically significant effect on ICP or CPP. There was a significant mean increase of 2.13mmHg in ICP from before to after oral care (p=0.0026). There was no significant relationship found between the ICP change and oral care duration (p=0.5687) or intensity (p=0.9154).

This study provides evidence that oral care, regardless of duration and intensity, is safe to perform in patients in the absence of pre-existing elevated ICP.

Critical care education in the surgical specialties has been experiential, fellowship based, and taught via the "see one, do one, teach one" method. This method does not incorporate a curricula-based approach to learning or patient care in the Intensive Care Unit (ICU). Simulation facilitates learning of skills without jeopardizing the safety of real patients. We developed a simulation model for Advanced Training in Neurocritical Care (NCC) in Neurosurgery (ATNS). Objective: To study the effect of simulation based training sessions on resident performance using an objective assessment of selected topics and principles in the management of neurologically critically-ill patients.

A total of 13 residents were separated into small working groups. Each group was given a specific case encountered in the Neuro-ICU. A NCC attending guided them through the case using a standardized tool. The "patient's" outcome was driven by the resident's knowledge, assessment and management skills. After completion, each case was reviewed with the residents. Each participant completed pre and post-tests and received educational materials on the topics covered. Topics tested for Module 1: Hypoglycemia/MI; Module 2: Sepsis/PEA arrest and Module 3: Delirium/Torsades. A passing score was 70% or greater.

The pre-training pass rate was 14%for Module 1, 0% for Module 2 and 29% for Module 3. The post-training pass rate was 57%, 100% and 100% for those same modules. There was no association between post graduate year and pass rate.

Simulation serves as a promising platform for the education of neurosurgery residents in NCC. We have successfully implemented a model of Advanced Training in Neurocritical Care (NCC) in Neurosurgery (ATNS). We will continue to explore and expand its use in residency education at our institution. 

The complex needs of neurosurgical patients present unique challenges across the continuum of care. The Trauma/Neurosurgery (TN) Program endeavored to improve patient access to care by planning for unexpected overnight admissions and enhancing the patient transfer process from the Neuro-Trauma Intensive Care Unit (TN ICU) to the TN Ward.

The TN program see 5 admissions overnight more than 50% of the time. In order to accommodate the Emergency patient, two less acute TN ward inpatients were bedspaced to alternate units prior to night shift to create two flow beds. The project also aimed to support an early morning patient transfer process between the TNICU and TN ward. Multidisciplinary brainstorming was held to identify processes required to discharge two patients by 09:00h from the TN ward, to be able to accommodate two patients by 10:00h from the TNICU. Process maps were developed to show the necessary communication, action items and transfer points that would support discharging patients earlier from the TN ward to accommodate TNICU transfers.

Flow beds were successfully created over 85% of time by proactively bedspacing patients within the hospital. This reduced overnight bedspacing of patients by 50% increasing staff and patient satisfaction. Emergency department metrics including length of stay and time from admission to bed were improved by 25% and 43% respectively at the 90 th percentile. Both changes were statistically significant at the 1% level. Total inpatient length of stay was reduced by over 10%.

Processes that focus on proactively planning for predictable demand, combined with strategies for demand and capacity matching, allowed for smoother flow through the TN program and an improved patient experience.

It has been suggested that 24/7 in-hospital coverage by licensed independent practitioners could improve outcomes following complex stroke including aneurysmal subarachnoid hemorrhage (aSAH). Due to a nationwide shortage of intensivists, such coverage is not easily acquired. Our objective was to compare outcomes in patients admitted with aSAH between two university-based tertiary referral centers with dedicated neurosciences critical care units (NCCU) and shared neurosurgical, endovascular and neurocritical care attending practitioners. NCCU-A is a 22-bed combined intensive care unit(ICU)-intermediate care unit (IMC) staffed 24/7, with overnight in-house NCCU fellow and resident coverage. NCCU-B is a 14-bed unit (8-bed ICU/ 6-bed flex IMC/ICU) covered overnight with home call by fellow and in-house resident covering the general ward and ED consults in addition to the ICU and IMC.

Prospectively collected data of aSAH patients admitted to the 2 institutional NCCUs between 2001-2009 were reviewed. We analyzed factors that impacted hospital SAH mortality, functional outcomes [dichotomized poor outcome: GOS 1-3] and related complications like aneurysm re-rupture and delayed cerebral ischemia (DCI). 

The non-inferior outcomes at the lower SAH volume center suggests that dedicated in-hospital 24/7 NCCU coverage may not be essential for delivery of high quality care in smaller sized neuroscience ICUs.

Data suggests improved outcomes in neurological critical illness if care is provided within a system of specialized Neurocritical Care (NCC). Given constraints in resources to deliver such care without a dedicated neurological ICU, we designed a novel 'virtual' NCC unit by implementing a new patient streaming system, a NCC team and standardized treatment protocols within an existing mixed ICU. This NCC service was envisioned as a rapid and cost-effective way to bring best practices to this patient population. In advance of its implementation, we assessed readiness and existing barriers among staff to these organizational and process changes.

The setting was a mixed Level III ICU, in an academic hospital and trauma centre in Toronto, Ontario, Canada. The qualitative study employed concurrent mixed methods, specifically stakeholder interviews, a staff survey and focus groups. Eighteen stakeholders were individually interviewed. 116 of 217 eligible ICU staff (53%) participated in the survey and 36 staff attended the focus group sessions.

From the survey, the most significant barriers to this re-organization were staff anxiety about coping (28%) and a concern that patients would not receive better care (24%). Noteworthy obstacles about the use of protocols were their lack of flexibility (19%) and that implementation was seen as impractical (16%). Seventeen barriers were proposed through an open-ended survey question. Content analysis revealed central barriers to be general resistance to change, educational challenges, workflow adjustment to a diagnosis-based rounding pattern and coordination conflicts. These findings were confirmed in focus group discussions, with a lack of resources as an additional important challenge.

When introducing a new system of providing Neurocritical Care, it would seem prudent to anticipate and resolve various barriers to success, which include lack of knowledge, work flow challenges and general resistance to change.

Severely elevated intracranial pressure (ICP) and clinically evident cerebral herniation are neurological emergencies and must be immediately managed to prevent substantial morbidity and mortality. Medical management relies on hyperosmolar therapy, which includes drugs that are difficult to stock everywhere in the hospital due to their high-risk nature, i.e. hypertonic saline. We have implemented a "Brain Code" system to accelerate administration of ICP emergency medications anywhere in the hospital. Creation of the system required approval from pharmacy, nursing, neurosurgery, neurocritical care, and hospital safety committees, among others, as well as education of hospital staff.

When an ICP emergency is recognized, the primary team pages a "Brain Code." This pages the code pharmacist (trained in Brain Code management), who delivers the secured Brain Code medication kit (containing 3% and 23.4% saline, mannitol 20%, phenylephrine 100 mcg/mL in a 10 mL syringe, and other supplies) to bedside. These medications are prepared by pharmacy and administered under protocol. A runsheet documenting code and medication administration times is completed by the pharmacist.

Brain Code runsheets were reviewed for code/medication administration times, and location. Patient charts were reviewed when runsheets were incomplete.

In the 16 months since its inception, the Brain Code has been activated 24 times. In 14 cases, Code times were documented. Average time from Brain Code activation to medication administration was 10.4 minutes. Brain Codes were called in the ER (52%), Neuro ICU (20%), Surgical ICU (9%), Helipad (4%), and PACU (4%).

The initiation of a "Brain Code" system facilitates expeditious administration of ICP emergency treatment, anywhere in a hospital. Further study is warranted to determine if such a system improves medication administration times over conventional systems, or patient outcomes.

In 2004, JCAHO found that a lack of communication resulted in 65% of all sentinel events and that 50% of those occurred during handoff. As we have begun the process of developing a new Neurosciences Critical Care Unit (NCCU), we wanted to assess our practice of RN handoff to identify specific areas for improvement. With those areas identified, we plan to develop a handoff template to improve the quality of care and patient outcomes.

We developed a 12 point questionnaire regarding the information an RN was given during handoff using a Likert scale with 6 possible responses. Our focus was on comprehensiveness in report that provides the necessary information to assume safe patient care. We queried 45 RN's over a 2 week period. Of those, 12 were NCCU RN's and 31 were SICU RN's that floated into the NCCU on a consistent basis.

Several key areas emerged from the data. Knowledge of the primary diagnosis, having the necessary information to assume responsibility for patient care, and knowledge of the plan of care were the areas of concern. The primary diagnosis was "always" known by the oncoming RN 69% of the time and "usually" known 31% of the time. Information to assume responsibility for patient care was "always" known 38% of the time and "usually" known 56% of the time.

Knowledge of the plan of care was "always" known 18% of the time, "usually" 62% of the time, and "occasionally" 18% of the time.

We are now able to create a handoff template that specifically focuses on the areas identified. In having a template, we can create a systematic way of giving and receiving handoff and the specific information that is necessary to convey.

Neurocritical care (NCC) is a subspecialty of critical care that is highly complex and demands a mastery of a deep understanding of brain physiology. NCC requires the provider have the ability to make rapid patient assessments, prioritize problems and anticipate the consequences of an action. As the number of NCC units increases across the country, the need for mid-level practitioners (MLPs) will also rise. However, specialty training does not readily exist and orientation programs frequently vary in quality without providing individuals with the tools they need to function at a competent level in NCC. This situation has led our team to develop a NCC Fellowship Program for MLPs.

This is a descriptive proposal on the formulation of the idea of a NCC fellowship program, development of the curriculum and implementation process. This is currently an ongoing prospective study of MLPs and students in their final year of training who are provided a year-long course focused on NCC. Methods for this research study include the development of a 5-module curriculum built around a concept using didactic lectures and simulations. All participants were given questionnaires and surveys on self-efficacy, job satisfcation and job prospects prior to beginning the course, which will be repeated at the completion of the program. The fellowship cohort will then be compared to their peers.

Preliminary results show an increased interest in working in a neurocritical care unit by student MLPs and increased involvement in NCC issues by the MLPs who are already working. In addition, the MLPs have become mentors to the students during simulation, thereby stimulating the student's interest.

The NCC fellowship program fills a gap in educational training for MLPs. The fellowship has the potential to stimulate interest in neurocritical care, recruit and retain highly educated MLPs and provide advanced specialty training in neurological diseases and sequelae.

Dexmedetomidine is associated with varying rates of hemodynamic derangements. Depending on dosing strategy and patient population, reported rates of hypotension have ranged from 23% to 98%, with bradycardia varying from 3% to 42%. Little is known regarding dexmedetomidine's adverse effect profile in neurocritical care, as these patients were excluded from most clinical trials regarding continuous sedation. We aimed to compare the incidence of severe hypotension and bradycardia in neurocritical care patients receiving continuous sedation with dexmedetomidine and propofol, the most commonly used sedative in this patient population.

This multi-center, retrospective study included neurocritical care patients between 18 and 89 years of age who received continuous sedation with dexmedetomidine or propofol for at least 4 hours. The primary outcome was a composite of severe hypotension, defined as a mean arterial pressure (MAP) <60 mmHg, or bradycardia (heart rate <50 beats per minute). Secondary outcomes included the individual incidence of hypotension and bradycardia, as well as the requirement for vasopressors to treat hemodynamic disturbances. A dexmedetomidine dosing protocol was utilized that withheld bolus doses and limited dose increases to every 30 minutes.

We included 342 patients in this study (105 dexmedetomidine and 237 propofol). There was no statistical difference in the incidence of the composite outcome between groups (29% vs. 30%, p=0.86). Analyzed separately, there were no statistical differences in the rates of hypotension (23% vs. 26%, p=0.52) or bradycardia (7.6% vs. 5.5%, p=0.45). Among patients who became hypotensive, there was no difference in median nadir MAP (56 vs. 56 mmHg, p=0.32) or the number of patients requiring vasopressors (15% vs. 25.9%; p=0.49).

With use of a dosing protocol, severe hypotension and bradycardia occur at similar rates in neurocritical care patients who receive either dexmedetomidine or propofol. Providers should evaluate the likelihood of hemodynamic instability before starting either sedative.

The devoted neurosurgical intensive care unit (NSICU) in rural area is paramount. Bangkok Trat hospital is at Thai-Cambodian border.The clinical teamwork in conjunction with teleconsultation is warranted for prompt response, enhancing the benefit to the patients.

A single center retrospective study was conducted during October 1 st , 2012 and March 31 st , 2013. The cases of intracranial hemorrhage (ICH) that had been consulted and transferred to the Bangkok Trat hospital NSICU (BTH-NSICU) were included. Admission demographic data, initial Glasgow coma score (GCS), time used for refer, GCS change during transfer, GCS at receiving hospital and outcome measurement according to Modified Rankin Scale (MRS) and Glasgow outcome score (GOS) at 1 month post injury were analyzed.

Eighteen cases were transferred to BTH-NSICU, traumatic ICH 12 cases (67%), non-traumatic ICH 6 cases (33%). Thirteen cases (72%) arrived in non-office hour, 4.00 pm-8.00 am. The transfer distances were between 5 and 240 kms. Average time used for refer was 154.28 min. They were operatively treated 14 cases (78%), non-operatively 4 cases (22%). There was no correlation between time used for referred and deteriorated GCS during transfer (r s =0.311, p=0.208, N=18). There were relation (p <0.05) between favorable outcome and initial GCS ≥8 (χ2 = 0.004), receiving GCS≥ 8 (χ2=0.018), motor (M) component ≥ 4 of initial GCS (χ2=0.023), M component ≥4 of arrival GCS (χ2=0.018). However, there was no relation between time used for refer ≥ 240 min and unfavorable outcome (χ2=0.317).

The devoted NSICU in rural area has much benefit. The well-organized team, workflow and effective teleconsultation help maximize the advantage and avoid unnecessary delay. The predictors of favorable outcome at 1 month post injury are initial GCS ≥8, arrival GCS ≥8, and the M component ≥4 of initial or arrival GCS. The RCSE recommended time using before surgery, 4 hours, may be lengthened if necessary.

To describe clinical diagnoses encountered during telestroke consultations and examine clinical differences between stroke and non-stroke diagnoses

This is a retrospective review from telestroke software. Data obtained between 7/2010 and 2/2011 from 8 hub and 24 spoke hospitals were abstracted for 533 consecutive consultations. Diagnoses were recorded and grouped as stroke (ischemic/TIA and hemorrhagic) and non-stroke (headache, medical diagnosis, specific neurological diagnosis other than stroke, and psychiatric). Basic demographic data, vascular risk factors, focal neurological exam findings and speech difficulties were recorded. Univariate analysis was used for statistical analysis.

Diagnoses were: stroke/TIA 220 (41.4%), intraparenchymal hemorrhage 19 (3.5%), headache 6 (1.1%), medical diagnosis 15 (2.8%), other defined neurological diagnosis 16 (3.0%), psychiatric 3 (0.5%), unspecified diagnosis 252 (47.4%). Those without the appropriate documentation where not included in the statistical analysis. Ischemic and hemorrhagic strokes (total patients 239) were grouped and compared to the remaining diagnosis (total patients 40). The prevalence of the above for the stroke group demonstrated: 96 males (43.6%), diabetes 64 (29.1%), hypertension 163 (68.2%), prior stroke 79 (33.0%), history of atrial fibrillation 29 (12.1%) and for nonstroke diagnosed patients: male 22 (55%), diabetes 9 (22.5%), hypertension 16 (40%), prior stroke 9 (22.5%), history of atrial fibrillation 5 (12.5%). Statically significant differences were the presence of hypertension (p=0.0006) and focal neurologic deficits (p=0.0016).

Telestroke consultations diagnosed multiple medical and neurological conditions. Hypertension was the only clinical feature that was statistically significantly different between the two groups. It was also shown the focal neurologic deficits documented by a neurologist was associated with stroke. At present there is not a recommendation for particular physicians completing telestroke consultations, those who do need to be prepared for a wide array of diagnoses and will need to be confident on tele-examination as this is one of the few results associated with both stroke and other neurological diagnosis.

Intracranial Pressure (ICP) is an icon of neurocritical care. Hundreds of recent studies have reported ICP as either a dependent or independent variable, but rarely do these manuscripts provide specific details associated with ICP measurement, and often these manuscripts describe multi-center studies. The untested assumption is that ICP monitoring and management practice is fairly standardized (adequately similar to be used as a research variable) across centers. The purpose of this study is to explore that assumption.

In the spring of 2013, the NCS members were emailed the link to a 14-item practice survey and asked to complete the survey anonymously (www.surveymonkey.com). The survey was open for two-weeks and a reminder (email) was sent after the first week.

The primary results from this survey, which reflects 241 practitioners who responded, have been published in the journal Neurocritical Care. For any given question, there was little agreement on best practice. From the series of six scenarios about treating ICP via CSF diversion, there was wide practice variation (category response rates ranged from 3.7% to 39.4%) and no clear preference (i.e., no category received > 50%). The highest agreement for ICP monitoring reference point was 48.4% (Tragus of ear). There were 50.2% who recorded values using only mm Hg (vs cmH20 or both mm Hg and cmH2O). A slim majority (52.3%) indicated that they record the displayed value (vs highest, mean, or most representative).

There is very little common-ground associated with ICP monitoring. The lack of consistent practice has broad implications for research and clinical practice. Studies which fail for control for practice variation will result in non-generalizable results and clinicians who are not provided the full details of inter-center practice preferences may initiate treatments that are not scientifically validated.

The purpose of this single-center study was to compare satisfaction with various aspects of care in our neuroscience intensive care unit (NICU) among family members of patients who were admitted with a neurosurgeon as the attending of record (Neurosurgery) with families of those admitted to an attending neurointensivist's primary service (Neurology).

From December 2012 to May 2013, a multidisciplinary team administered the previously validated 24-item Family Satisfaction-ICU (FS-ICU) survey to a consecutive cohort of families at time of NICU discharge. Only those families whose relatives were admitted for less than 72 hours or whose relatives passed away during NICU admission were excluded.

Of 121 eligible patient families, a total of 67 families (55.4%) had one member complete the survey (30 Neurosurgery, 37 Neurology). When asked in particular how well the NICU staff provided emotional support, 76.6% of neurosurgery families indicated complete satisfaction with care, compared to 51.4% of Neurology families (p = 0.04). Neurosurgery patients on average were younger than neurology patients (57.2 versus 69.0 years, p=0.009) and had a longer mean length of NICU stay (13.4 versus 6.4 days, p=0.0001). The most common neurosurgery diagnoses were subarachnoid hemorrhage (50.0%), traumatic brain injury (10.0%), and subdural hematoma (10.0%). Intracerebral hemorrhage (37.8%), ischemic stroke (24.3%), and seizure (10.8%) were the most frequent neurology diagnoses. Neurology patients were more likely to be privately insured (54.1% versus 40%, p=0.004), while neurosurgery patients were more likely to be covered under Medicaid (33.3% versus 10.8%, p=0.03).

In our NICU, neurosurgery families are more likely to be completely satisfied with the emotional support received from staff compared to neurology families. Whether this difference is intrinsic to the different patient and family populations triaged to the two services or represents a systemic, correctable practice difference should be explored.

Limited resources , neurointensivists, and neurologic intensive care unit (neuroICU) beds warrant investigating models for predicting who will benefit from admission to neuroICU. This study presents a possible model for identifying patients who might be too well to benefit from admission to a neuroICU.

We retrospectively identified all patients admitted to our 16 bed Neuro-ICU between 11/2009 and 2/2013. We used the APACHE Outcomes database to identify patients who on day one of neuroICU admission received one or more of 33 subsequent active life supporting treatments. We compared two groups of patients: Low Risk Monitor (LRM)(Patients who did not receive active treatment on the first day and whose risk of ever receiving active treatment was <10 %) and Active Treatment (AT)(patients who received at least 1 of the 33 ICU treatments on any day of their ICU admission).

There were 873 admissions (46%) in the LRM group, and 1006 admissions (54%) in the AT group. ICU length of stay (LOS) in days was 1.7 (+/-1.9) for the LRM group versus 4.5 (+/-5.5) for the AT group. ICU mortality was 0.8 % for the LRM group compared to 14 % for the AT group (OR = 17.6; 95% confidence interval [CI], 8.2 -37.8, p <0.0001). Hospital mortality was 1.9 % for the LRM group compared to 19 % for the AT group ( OR = 9.7; 95% CI, 5.8 -16.1, p <0.0001).

The outcome for low-risk monitor patients in our neuroICU suggests they may not require admission to neurologic intensive care. This may provide a measure of neuroICU resource use. Improved resource use and reduced costs might be achieved by strategies to provide care for these patients on floors or intermediate care units. This model will need to be validated in other neuroICUs and prospectively studied before it can be adopted for triaging admissions to neuoICUs.

The UC San Diego hospital system consists of two main hospitals, Hillcrest and Thornton. A primary Neurocritical Care (NCC) service, under the direction of an NCC fellowship-trained/board-certified neurointensivist, was started in January 2012 at Hillcrest only. An endovascular neurosurgeon was hired to operate at both facilities. The NCC service consists of an NCC team (three NCC-trained attendings with continuous resident coverage) that admits and provides primary management for all ICU-level neurosurgical/neurological patients. The majority of NCC patients are housed in one ICU with nurses NCC-educated by the NCC director.

We analyzed United Health Consortium (UHC) data for all NCC patients from July 2012 -March 2013, comparing with Hillcrest neurosurgical/neurological ICU patients from July 2010 -July 2011, prior to NCC establishment. We also analyzed data on neurosurgical/neurological ICU patients treated at Thornton. These patients are co-managed by Neurology/Neurosurgery ward teams with Surgical or Pulmonary Critical Care services. Current annual case volume was extrapolated from the July 2012-March 2013 case volume, in order to compare with annualized data from 2010-11.

At Hillcrest, we found an increase in NCC cases/year (+126, +51%), decrease in ICU length of stay (LOS) (-2 days, -33%), decreased observed/expected (O/E) in-hospital LOS (-0.23, -17%), decreased O/E in-hospital mortality (-0.22, -22%), and decreased O/E direct cost (-0.03, -2%). At Thornton, although NCC case numbers also increased (+84, +75%), there were increases in average ICU LOS (+1.2 days, +48%), O/E in-hospital LOS (+0.23, +20%), O/E in-hospital mortality (+0.65, +59%) and O/E direct cost (+0.39, +33%).

Implementation of a primary NCC team benefited patients and the hospital by increasing case volume, while decreasing ICU LOS, O/E LOS, O/E mortality, and direct costs. These preliminary results validate the initial success of the program and suggest impetus for its expansion.

Financial Support: None S176 Neurocrit Care (2013) 19:S1-S331 

Critically ill patients receiving prolonged sedoanalgesia (SA) for adaptation to mechanical ventilation (MV) are at risk for developing neurological complications while at the same time their neurological evaluation is limited. We describe the pupillary function in patients without neurological or ocular injury that were exposed to MV for respiratory insufficiency and received either Remifentanil or Dexmedetomidine for SA.

A prospective cohort study was performed in a general ICU at a large academic medical center. We studied pupillary function in 27 patients using a pupillometer Neuroptics ® . A total of 6 measurements were performed in each individual: 4 while receiving sedoanalgesia (OnSA) and 2 after discontinuation of it (OffSA) with patients fully awake. During the OnSA measurements all patients had the same depth of sedation.

In the Remifentanil group (n: 11) the mean pupillary diameter was 2.91 mm (OnSA) and 4.22 mm (OffSA) (p=0.005). After exposure to light, the percentage of change of pupillary diameter was 12.07% (p=0.0009) and both the Maximum Constriction Velocity (MCV) and Dilatation Velocity (DV) changed significantly when comparing OnSA versus OffSA. In the Dexmedetomidine group (n: 16) the mean pupillary diameter was 3.12 mm (OnSA) and 3.83 mm (OffSA) (p=0.011). After exposure to light, the percentage of change of pupillary diameter was 6.14% (p=0.001) and both the MCV and DV changed significantly when comparing OnSA versus OffSA. The Latency Period did not change in any group.

Dexmedetomidine and Remifentanil both alter pupillary physiology causing miosis and decreasing the constriction and dilatation velocities. None of the studied drugs affects the Latency Period suggesting their lack of action on the afferent pathway of the light reflex. Knowing the "expected normal" range of pupillary diameter in patients under SA with the studied drugs could help detect acute systemic or neurologic complications that may manifest as changes in the pupillary diameter.

Performing routine and non-routine bedside portable radiographs is an important part of daily ICU workflow for patients that are intubated and mechanically ventilated. The aim of this study was to determine the incidence of performance of imaging and abnormalities noted on routine and non-routine chest x-rays in critically ill patients.

This retrospective chart review was conducted on patients admitted to the neuroscience intensive care unit of a major academic medical center between January 1st, 2009 to March 31st, 2012. Only those patients who were endotracheally intubated and mechanically ventilated were selected for this study

The median age of our cohort was 61 years. A total of 859 chest x-rays (CXR) were ordered. 67% of these were ordered as routine, while 33% were ordered as non-routine. Mean number of routine CXR per patient per admission day was 0.56. Average number of non-routine CXR per patient per admission day was 0.28. Majority of chest x-rays performed in any given patient (74%) were done during the phase of endotracheal intubation. Of 578 routine CXR, 65% were abnormal but stable, while only 25.26% CXR were interpreted as being abnormal and unstable. 95% of all post-intubation CXR performed demonstrated optimal endotracheal intubation depth. 98% of all post central venous catheter CXR performed demonstrated optimal central venous catheter placement and absence of any complications. When CXR were ordered for following up on any existing abnormality, 59% CXR were interpreted as stable.

In our cohort, only a minority of intubated and mechanically ventilated patients demonstrated a new abnormality. Only 52% of CXR ordered for evaluation of change in clinical condition had a new abnormality . Routine chest x-rays failed to demonstrate a significant number of endotracheal tube malpositioning. The practice of routine and non-routine imaging in the ICU must balance physician bias, costs and hazards of radiation exposure.

The presence of cerebral microhemorrhages(CM) has been associated with intracranial hemorrhages in patients with cerebral amyloid angiopathy and cardiovascular disease risk factors. Leukemia is complicated by intracranial hemorrhages with grave consequences. We hypothesized that the presence of microhemorrhages in patients with leukemia will be associated with intracranial hemorrhages.

We retrospectively reviewed our database of adult (>18 yrs) patients with leukemia who had a magnetic resonance imaging (MRI) with susceptibility weighted imaging(SWI) or gradient echo(GRE) sequences done for neurological symptoms during their clinical course of leukemia diagnosis and management. Demographics, diagnosis, history of hypertension, diabetes mellitus(DM), coronary artery disease(CAD), congestive heart failure (CHF), presence of intracranial hemorrhages-subdural(SDH), subarachnoid(SAH) and intracerebral hemorrhages(ICH) and laboratory dataplatelet count (x 10 3 /mm 3 ), international normalized ratio (INR, secs), partial thromboplastin time (PTT, secs) and serum creatinine at time of MRI were obtained. Statistical analysis was done using fisher exact test for categorical variables and t-test for continuous variables.

Of 87 

In leukemia patients with neurological symptoms, the presence of CM may be associated with increased risk of intracranial hemorrhages and may serve as a marker for intracranial hemorrhagic complications.

Serotonin syndrome (SS) is potentially becoming more frequently encountered in the intensive care unit (ICU). We sought to determine the clinical presentation, drug exposures, and outcomes of SS in critically ill patients.

A retrospective study of 33 consecutive patients with SS between 2007 and 2012 in ICUs in a large teaching hospital. SS was defined using the Hunter Serotonin Toxicity Criteria.

Seventeen patients (50%) were admitted for mental status changes, but SS was the primary admission diagnosis in only 4 (12%) cases. The rest were admitted for various other reasons and in 13 patients (39%), the features of SS developed only after a mean of 6.8 ± 9 days of hospitalization. Most received multiple serotonergic drugs upon diagnosis (median 3 drugs, range 1-5). Antidepressants were the serotonergic medications most often used before admission, and opioids (principally fentanyl) and antiemetics were the most frequently prescribed new serotonin-enhancing medications. Altered mental status was present in all patients and myoclonus, rigidity, and hyperreflexia were the most prevalent examination signs. All but one patient had documented recovery. The mean time to neurological improvement was 56 ± 5 hours, but ranged from 8 to 288 hours. There were no cases of renal failure related to rhabdomyolysis, or death or persistent disability caused by SS.

SS in the ICU occurs most often because of exposure to multiple serotonergic agents. Continuation of antidepressants plus the addition of opioids and antiemetics during hospitalization are most commonly responsible for this complication.

Status epilepticus following cardiac arrest (SECA) has been associated with poor clinical outcomes. Recent studies suggest this association may not be true. We sought to test the hypothesis that SECA does not predict higher in-hospital case fatality.

We conducted a case-control study using data from United States ICU's from 2003-2008. Out of 4160 patients with cardiac arrest (CA), we identified 16 patients with a secondary diagnosis of status epilepticus (SE). We randomly selected 32 control patients with CA without SE.

A total of 16 case and 32 control patients were analyzed. The mean age was 54 yrs. (SD16) for case subjects, 64 yrs. (SD17) for controls (p=0.05), female cases 7 (44%), female controls 17 (53%), Caucasian cases 6 (40%), controls 23(80%). Case fatality rates are similar for SECA cases and CA without SE controls (11/16=68%) vs. (20/32=63%) (p=0.6). Preadmission functional status: independent 12 (80%) cases vs. 25(78%) controls, partially dependent 1(7%) case vs. 5 (16%) controls, fully dependent 2 (13%) cases vs. 2(6%) controls. GCS (Md, IQR): cases 3 (3-3), controls 3 (3-8) (p=0.03). SAP-II Mean (SD): 13(65) cases vs. 29(62) controls (p=0.6), Organ Dysfunction in the ICU N%: 12(75%) cases vs. 18 (56%) controls (p=0.2).

Our results suggest that SECA is not associated with higher in-hospital case fatality. This finding supports research suggesting that status epilepticus following cardiac arrest is not a predictor of higher mortality, particularly in the era of hypothermia. Our findings require further investigation.

Fulminant hepatic encephalopathy (FHE) has dismal prognosis if transplantation is not an option. Novel therapeutic interventions may change this outcome.

We reviewed all cases with FHE admitted to our hospital since 2008. In 2010 we developed a multidisciplinary management protocol for patients with this condition admitted to our hospital. This mandated transfer of patients entering grade 3 from other Intensive Care Units (ICUs) to the Neurosciences-ICU (NICU) for intracranial pressure (ICP) management. Multiple interventions were utilized including coagulopathy reversal with Factor VII and prothrombin complex concentrate (PCC), ICP device placement, hypertonic osmotic treatment, aggressive ammonia lowering regimen with lactulose and rifaximin, early renal replacement therapy, mild hypothermia for refractory ICP, all in parallel with liver transplantation candidacy investigation. Only patients admitted to the NICU received FVII/PCC and a structured management protocol.

Twelve patients (11 women, mean age of all patients 39 years) were admitted; four were managed in the MICU/SICU (2/4 before 2010) and 8 in the NICU (all after 2010). The etiology of FHE was acetaminophen toxicity in 9/12 patients. The Model for End-Stage Liver Disease (MELD) admission scores, ammonia levels and liver enzymes between the MICU/SICU and the NICU were not different (Mann-Whitney test). Ten patients received ICP monitoring (all in the NICU plus 2 in the SICU) and the highest ICP recorded was 111 mm Hg. The pre-and post-coagulation reversal INR were 3.5 and 1.3, p=0.01, Wilcoxon test). Three patients in the NICU received hypothermic treatment. Five patients died, with mortality in the MICU/SICU being 75% (3/4) and in the NICU 25% (2/8). Two patients were transplanted new liver, one of whom later died.

A multidisciplinary approach and development of a protocol-driven management tool based on novel interventions may improve the outcome of patients with FHE.

Mechanical ventilation in patients treated with mild therapeutic hypothermia (MTH) for post cardiac arrest syndrome may affect cerebral blood flow and acid-base balance, but the optimal management strategy remains unknown. Alternative strategies depend on whether or not arterial blood gas (ABG) measurements are temperature-corrected (TC). We sought to describe practice variability in TC at a single institution and used such practice variability to explore the association between TC and outcome.

We conducted a retrospective cohort study by reviewing electronic health records of all patients treated with MTH after cardiac arrest. We reviewed all ABGs drawn during hypothermia therapy. We tested whether the percentage of TC ABGs relative to total number of ABGs drawn for each patient was associated with neurological outcome (dichotomized cerebral performance category, CPC) via logistic regressions.

1259 ABGs were obtained during MTH on 122 eligible patients over 6 years, of which 362 (29%) were TC. TC was never used in 72 patients (59%), and was used in ≥75% of ABG's in 33 (27%). The latter group had a higher proportion of patients with shockable presenting rhythms (42% vs 76%, p=0.0012), but the two groups were otherwise similar. Favorable outcomes at approximately 6-month follow-up were more common in those with ≥75% TC ABGs (15/31, 48%), than in those with no-TC (16/65, 25%), but this relationship was not significant after adjusting for presenting rhythm (OR 1.7, 95% CI 0.62-4.5). Practice variability appeared driven by provider rather than patient factors.

There was substantial practice variability. Ventilator management based upon TC of ABG's versus no correction was not associated with robustly better outcomes in patients treated with MTH after cardiac arrest after adjusting for presenting rhythm . Choosing whether to TC may not be clinically important in this population. 

Acute encephalopathy in critically ill patients is common and is associated with high mortality. Preservation of physiologic sleep has been associated with favorable outcomes in acute brain injury. We hypothesize that presence of elements of sleep architecture in EEG is associated with better outcome in adult patients with acute encephalopathy.

Clinical data, EEG characteristics, and outcome were assessed in 142 consecutive patients with EEG obtained for acute change in mental status. EEGs were interpreted regarding presence of sleep elements (i.e., K-complexes, vertex sharp waves, and sleep spindles). Associations between sleep elements and outcome measured by Glasgow Outcome Scale (GOS) were assessed.

142 patients were included: 41.5% male, median age 64.5 (range 18-98). Leading etiologies were infections (47.2%), intracranial hemorrhages (14.1%), and acute ischemic strokes (10.6%). There were no differences in medical comorbidities between groups. Of the EEGs obtained at time of mental status change, all demonstrated patterns of encephalopathy and 38% had at least one sleep element. 27.5% had K-complexes, 31.7% had vertex waves, and 33.8% had sleep spindles. Patients without sleep elements were older (median age 68.5 vs. 58, p=0.01) and septic shock was more common (10 vs. 0, p=0.014). Sleep elements were not associated with mortality or outcome; however, even when adjusted for confounders (age and septic shock), the presence of K-complexes was significantly and independently associated with good short-term outcome (GOS5, multivariable logistical regression, p=0.022).

The presence of K-complexes on EEG obtained for alteration in mental status was a robust association with good outcome, but absence of sleep elements did not predict poor outcome. While these findings need to be confirmed in larger prospective studies, it gives us critical insight on the preservation of sleep elements as a marker for favorable outcome in patients with acute encephalopathy.

Weight-based prothrombin complex concentrate (PCC) is used to rapidly reverse warfarin-related coagulopathy in patients with acute intracranial hemorrhage. However, not all patients achieve full reversal of international normalized ratio (INR) after the first calculated dose. We sought to identify clinical factors that are associated with anticoagulation reversal failure after the first calculated dose of PCC.

We performed a retrospective study of consecutive patients with acute intracranial hemorrhage and warfarin-related coagulopathy who were hospitalized at a tertiary center between 1/1/2010 and 12/31/2012. Anticoagulation reversal failure was defined as INR ≥ 1.5 after the first calculated dose of PCC. Multivariable analysis using logistic regression was used to determine the predictors of anticoagulation reversal failure after the first calculated dose of PCC.

During the study period, 51 patients with acute intracranial hemorrhage received PCC for rapid warfarin reversal using a simple weight based protocol. Overall, 23 out of 51 (45%) patients did not achieve full reversal of INR. Those with anticoagulation reversal failure were obese (BMI > 30kg/m 2 ) (41% vs. 14%, p = 0.03), had a higher initial INR (3.0 ± 1.4 vs. 12.0 ± 0.7, p = 0.001), and had an initial INR >2 (22% vs. 67%, p = 0.001), compared to those who were successfully reversed after the first dose. In the multivariable logistic regression model, the independent predictors of anticoagulation reversal failure were obesity (odds ratio 7.88, 95% CI 1.12 to 55.68) and initial INR >2.0 (odds ratio 12.49, 95% CI 2.27 to 68.87).

A great proportion of patients were unable to be successfully reversed to an INR<1.5 after the first calculated dose of PCC, using the simple weight based protocol. The independent predictors of failure were obesity and initial INR >2.0. An individualized protocol with higher PCC dose may be needed for this patient population.

The exposure to neurocritical care among neurology residents in the country is highly variable. Recent advances in computer technology have enabled innovative medical education programs to incorporate high-fidelity simulation based learning (SBL) using mannequin as an instruction tool for teaching neurological emergencies to neurology residents. We hypothesize that it's feasible to teach neurological emergencies using SBL.

Eligible neurology residents (post-graduate year 2) from Columbia university and Weill-Cornell neurology programs (N=20) who have completed a year of medical internship were randomized into two groups-SBL and traditional didactic teaching. High-fidelity Sim-Man 3G was used to simulate realistic scenarios. Topics included status epilepticus, acute ischemic stroke and intracranial pressure crisis. Learning objectives will be assessed using crisis resource management (CRM) assessment tools including identification of key actions, time to key actions, Ottawa global rating score and Ottawa CRM checklist, and finally knowledge basedtests both pre-and post-intervention. Learner satisfaction through survey feedback and learner retention will be assessed on the same outcomes for both groups at 3 months.

Participants were effectively randomized to either SBL or didactics for each of the three cases. It was feasible to simulate real life scenarios for common neurological emergencies present in a neuro ICU setting. The analysis comparing SBL with traditional didactic learning for neurological emergencies as measured by knowledge based test and CRM assessment tools described above are pending.

Simulation based learning offers promise as a tool for objectively assessing some of the ACGME competencies that are more difficult to evaluate via traditional means. Although SBL can only provide an approximation of real-life resuscitation, it remains the only viable means to formally evaluate performance of critical care skills in real-time and dynamic environment and has great potential to become a validated and essential component of neurology residency training moving forward.

Financial Support: This abstract has been supported by the American Academy if Neurology Education grant 2013 and the Apgar grant 2013.

While numerous cases of orthostatic coma (OC) have been reported under various names, such as "brain sag," "paradoxical herniation," "intracranial hypotension" or "syndrome of the trephined", the sporadic nature of these reports leaves the condition poorly characterized. Though typically responsive to appropriate therapy, OC can be fatal when such therapy is not rapidly applied. Moreover, though OC may clinically mimic typical herniation syndromes, the pathophysiology and treatments are quite different. With this in mind, we performed a systematic review of existing OC case reports in order to better characterize this syndrome.

Authors independently searched the PubMed database for case reports or series describing OC. Data points regarding clinical scenario, presenting symptoms, neurologic exam, imaging, treatment, and response were extracted from identified case reports and analyzed qualitatively and quantitatively.

Forty-five cases were identified from 25 publications. Craniotomy was the most common inciting clinical scenario (n=27), followed by craniectomy (n=10), CSF leak (n=5), and spinal instrumentation (n=3). All patients who received craniotomy had a documented cause for significant CSF loss during or following their surgery that preceded the onset of OC, though exact volumes were infrequently reported. Eighty percent of post-craniectomy patients had lumbar puncture prior to cranioplasty that precipitated OC. Overall, no single clinical or imaging finding suggestive intracranial hypotension was reliably seen in all inciting scenarios. Postcraniectomy patients demonstrated contralateral midline shift on imaging, while subdural fluid collections were frequently present in other etiologies.

OC is a potentially fatal condition that is frequently unrecognized or misdiagnosed as herniation from intracranial hypertension. No single imaging or clinical finding can reliably differentiate these pathologies. However, cranial or spinal invasion combined with significant CSF loss reliably precede OC, and these scenarios must be avoided whenever possible. In patients with this clinical history and rapid neurologic decline, OC should be suspected and decisively managed.

To determine the relationship between enteral nutrition (EN) interruptions and calorie intake; calorie intake and clinical outcomes; and EN interruptions and outcomes among Neuro-ICU patients.

A prospective cohort study in mechanically ventilated (MV) brain-injured patients in an academic Neuro-ICU. The following outcomes were ascertained: MV days, ICU and hospital length of stay (LOS), rates of aspiration pneumonia (PNA), and in-hospital mortality. Patients were divided into two exposure groups based on receiving >60% (Group-1) or ≤ 60% (Group-2) of prescribed EN calories according to established criteria.

Total of 27 patients, of whom 26% had ischemic stroke, 7% had SAH, 44% had ICH, 15% had brain-tumor, and 7% had other diagnoses. Median(Md) age was 69yrs(IQR,63-79), 52% males, Md GCS=8(IQR,7-14), and Md BMI=28kg/m 2 (IQR,26-32). There were 20%(7/27) patients in Group-1 and 80%(20/27) in Group-2. Group-1 had more EN interruptions (mean 4.1; SD 0.9) than Group-2 (mean 2.8; SD 1.5)(p=0.01). Md time for feeding tube placement was 4.5hr in Group-1 (IQR,2-7) and 7hr in Group-2 (IQR,0-26)(p= 0.8). Md time to start EN in Group-1 was 17hr (IQR,13-29) and 28hr (IQR,23-61) in Group-2(p=0.05). There were no differences regarding duration of MV (p=0.09), ICU-LOS (p=0.1), hospital-LOS (p=0.09), and in-hospital mortality (p=1.0). The incidence of PNA was 26%(2/7) in Group-1 and 0%(0/27) in Group-2 (p=0.01).

In our Neuro-ICU, MV brain-injured patients were likely to be underfed. This phenomenon could be attributed to longer time to start EN rather than on frequent interruptions. As expected, EN interruptions were more likely in those with faster and higher rates of feeding tube insertion. Patients with goal caloric intake experienced higher rates of PNA, but this did not have a significant impact on other outcomes. We acknowledge that we may have been underpowered to detect meaningful differences. A larger-scale quality improvement project will help clarify our preliminary results.

Patients with acute brain injury (ABI) frequently require continuous infusions of analgesics or sedatives for discomfort or control of elevated intracranial pressure (ICP). However, patients with ABI also require frequent clinical exams. Success in maintaining patient comfort, physiologic goals, and performing frequent examinations can be a daunting task.

Dexmedetomidine, an α-2 adrenoreceptor agonist, has unique sedative and analgesic properties which may help make this task easier. However, little is known about its effect on cerebral perfusion pressure (CPP) variability. The purpose of this study is to determine the effects of Dexmedetomidine as an add-on therapy to current standard-of-care (SOC) sedation practice while exploring the effects on cerebral perfusion parameters.

This is a randomized, double-blinded, placebo-controlled trial of Dexmedetomidine as an adjunctive sedative in ABI patients. After obtaining approval from the Duke IRB, adult patients requiring ICP monitoring and continuous sedative or analgesic administration were eligible. Enrolled patients were randomized to receive SOC + Dexmedetomidine or placebo for 24 hours. Various physiologic parameters and medication infusion rates were collected continuously for the study using ICPilot tm software.

This study remains open for enrollment with 77 patients of various ABI types including subarachnoid hemorrhage, intraparenchymal hemorrhage, and post-operative cerebral edema. 66% have been Caucasian and 36% have been female. Of note, there have been no serious adverse events and only 1 patient was taken off of study drug due to bradycardia. However, all study drug assignments remain blinded at this point and are under intent to treat.

We present this abstract of an important study currently in progress. Lessons learned during the conduct of this trial will be presented as final results of the primary hypothesis are pending. However, subject recruitment and challenges of randomizing subjects to a study drug approved for regular use will be discussed as they influence clinical trial design.

Financial Support: Hospira, Inc.

The process of measuring intracranial pressure (ICP) can be accomplished using a variety of monitors placed primarily either in the ventricles or brain parenchyma. There is inadequate data to support the conclusion that ICP measured simultaneously from two different sites using two different devices produces similar findings in the same subject.

Retrospective observational study. Manual chart abstraction was used to obtain time-indexed ICP values from patients admitted between October 2010 and October 2012 and diagnosed with severe traumatic brain injury from the Parkland Hospital Trauma Databank. Indications for monitors were by Brain Trauma Foundation Guidelines. Only patients with simultaneous EVD and intraparenchymal (IPM) Camino placement and time-indexed charted data were considered eligible for inclusion.

Subjects (N=37) were primarily male (81%) with a mean age of 30.8 years (range 13-62 years). There were 2,547 timeindexed data points (ICP recorded from EVD and IPM). Patients were monitored with both devices for a mean of 71 hours (range 12-216 hours). Only 4 of 37 (10.8%) patients had strong (>0.40) or very strong (>0.70) correlations given they met the criteria from paired T-test. Within-subject Pearson correlations ranged from -0.43 to +0.99. The majority of withinsubject paired T-tests (N=29) found significantly different (p<0.001) ICP values recorded by EVD and IPM. ICP values obtained from the IPM exceeded those from the EVD 69% of the time by a mean difference of 4.5 mmHg.

Simultaneously recorded ICP values show a wide range and inconsistent direction of correlations. There was a small percent of subjects with strong or very strong correlations, and a high percentage of statistically significant different paired observations. The reason for the inconsistent correlations is unknown. There is inadequate evidence to support that intraparenchymal ICP values can be treated in a similar manner to ICP values obtained from an EVD.

Residents care for critically ill patients and have frequent bedside interactions with families that may influence decisions on medical care. We compared residents' predictions to patients' 6-month outcomes.

We prospectively enrolled consecutive neurological patients that required mechanical ventilation for ≥ 72 hours. Anesthesia, medicine, and emergency medicine residents caring for each patient were asked to predict: 1) 6-month modified Rankin scale (mRS), dichotomized into good (mRS 0-3) and poor (mRS 4-6) outcome, 2) 6-month quality of life (QOL; dead/NA, poor, fair, good, or excellent), and 3) whether care should be withdrawn. Six-month outcome was determined by standardized telephone interviews.

Of 347 eligible patients, we enrolled 272. Three were lost to follow-up, one withdrew, and residents failed to complete 60/268 of the questionnaires. In the final cohort of 208 patients, common diagnoses were stroke and status epilepticus. Survival was 54% (111/208). Overall, residents correctly predicted 6-month functional outcome (good vs. poor mRS) in 77% (95% CI, 71-83) of patients. However, 42% (28/67) of survivors who went on to have good outcomes were incorrectly predicted to do poorly. The residents' incorrect predictions were significantly more pessimistic than optimistic (28 vs. 12, p = 0.017) after excluding patients who had life support withdrawn. When residents predicted good/excellent QOL (n=18) they were correct in 61% (95% CI, 36-82) of cases and for poor/fair QOL (n=67), 49% (95% CI, 37-62) of cases. Withdrawal of care was recommended in 51 patients. Of these, 11 survived with a 6-month mRS 4-5, but two had mRS 1 and 3.

Residents have moderate overall accuracy in predicting functional outcome among neuro-ICU patients that require mechanical ventilation for ≥ 72 hours. Importantly, 42% of survivors with good outcomes were predicted to do poorly by residents. Further studies are warranted to assess if resident prognostication affects medical decision-making in the ICU.

Despite high rates of venous thromboembolism (VTE) among critically ill patients with intracranial hemorrhage, there is a reluctance to use pharmacologic VTE prophylaxis in patients with external ventricular drains (EVD) due to concerns of procedure site hemorrhage.

In a prospective study of intracranial hemorrhage patients with an EVD (subarachnoid hemorrhage N=64; intracerebral hemorrhage N=41 and subdural hemorrhage N=5) conducted between 7/2008-11/2011, we compared bleeding complications, VTE rates and 3-month functional outcomes between patients who received pharmacologic VTE prophylaxis (either heparin or enoxaparin) plus compression boots versus mechanical prophylaxis (compression boots) alone.

Of 110 patients with an EVD, 98 (89%) received pharmacologic prophylaxis (heparin BID N=3 [3%]; heparin TID N= 34 [31%]; enoxaparin N=61 [56%]) in addition to compression boots, while 11 (11%) received compression boots alone. The median time to initiation of pharmacological prophylaxis was 4 days post hemorrhage. Compared to mechanical prophylaxis alone, pharmacologic prophylaxis patients were more likely to have SAH (60% versus 22%, P=0.028), less likely to have SDH (2% versus 33%, P<0.0001) and less likely to be DNR/comfort care (19% versus 44%, P=0.035). Admission GCS and APACHE 2 scores did not differ between groups. There were no significant differences in bleeding events (EVD associated hemorrhage, ICH expansion, new ICH, new SDH, SDH reaccumulation, bleeding at the craniotomy site, GI bleeding or anemia requiring transfusion) or thombotic events (DVT or PE) between groups. At 3months, fewer patients receiving pharmacologic prophylaxis were dead (25% versus 67%, P=0.012) or severely disabled (51% versus 89%, P=0.049). After adjusting for admission GCS, age, DNR status, and bleed type this trend for reduced 3month mortality remained (adjust OR 0.1, 95%CI 0.004-1.2,P=0.065).

In intracranial hemorrhage patients with EVDs, pharmacological VTE prophylaxis is not associated with higher bleeding risks and is associated with improved 3-month mortality rates compared to mechanical VTE prophylaxis alone.

Ethics consultations are common in the intensive care unit, but there is very limited data regarding the reasons for consultations in the neuroscience intensive care unit (NICU) beyond issues regarding withdrawal of care. We wish to describe the spectrum of reasons for ethical consultations in a NICU within an academic medical center.

Through medical record review, we prospectively collected ethics consultations for NICU patients hospitalized in a single institution from 2006 through 2012 and categorized the reasons for the consultations. Our institution is the Brigham and Women's Hospital, a 793 bed, level 1 trauma center with a 20 bed NICU located in Boston, Massachusetts.

Over a 6-year period, there were 142 ethics consults performed on patients in the NICU. The categorical breakdown of reasons for consults includes: Proxy Decision Making: 46 (32%), Relationship Conflict/Treatment Futility: 32 (23%), Withdrawal/Withholding Support: 30 (21%), Communication: 19 (13%), Discharge Planning: 7 (5%),Decisional Capacity: 6 (4%), and other 2 (1%)

We present some of the first data to show the categorization of the reasons for ethics consultations in a NICU. In this single center study, we found the most common reason for an ethics consultation was to establish proxy decision-making given that virtually all patients in the NICU have sustained brain injuries resulting in a range of decision making incapacity. This information is helpful in guiding healthcare professionals to plan meetings regarding goals of care.

Over the past decade, there has been an increased use of continuous electroencephalography (cEEG) for non-invasive monitoring of cerebral function in critically ill patients. Whether cEEG results impact clinical decision making is not clear. The purpose of this study was to review the utilization of cEEG in our ICU, and to address whether or not clinical decisions were made on the basis of cEEG results

This study was a retrospective review of cEEG done in a 33 bed medical/surgical ICU from January 2008 -June 2013.

52 cEEGs were performed during the study period. The most common indication for cEEG was to rule out non-convulsive seizure or status epilepticus (51.9%; n=27). Within this group, epileptiform activity was detected in 55.6% of patients and its presence led to a change in clinical management in almost all patients (93.3%). Other indications for cEEG included monitoring during hypothermia after cardiac arrest (15.4%, n=8), clinical activity suspicious for seizure (21.2%; n=11), and unexplained decreased LOC (9.6% n=5). In patients who displayed clinical activity suspicious for seizure, an event was captured in 73% of patients. Interestingly, none of these events correlated with electrographic epileptiform activity. This absence of epileptiform activity led to a change in management in 75% of patients, usually weaning of sedation and/or anticonvulsant medications. Among all patients who underwent cEEG and ultimately had withdrawal of life support, cEEG data was used as part of the decision process in 32%. Common findings in these cEEG were epileptiform activity, burst suppression pattern, and lack of reactivity.

In our centre, cEEG appears to contribute to clinical decision making, including the titration of sedation and anticonvulsant medications, as well as withdrawal of life support. Larger, prospective studies are required to validate these findings.

The primary goal of this study was to quantify the differences in mean arterial pressure (MAP) measured by two techniques; noninvasive oscillometric (cuff), and indwelling arterial line (a-line)

A prospective observational study of 174 consecutive patients admitted to the Neuroscience Intensive Care Unit was conducted..With the patients at 30 degrees head of bed elevation, blood pressures were measured by cuff and a-line, with the transducer leveled at phlebostatic axis and the external auditory meatus.

Median age of our cohort was 58 years. The mean BMI was 29.32 kg/m2. The MAP measured by a-line leveled at the phlebostatic axis was higher than that measured by cuff (mean 8.6 mmHg, standard deviation 10.7 mmHg,p-value <0.001, 95% CI 7 -10.

2) The MAP measured by a-line leveled at phlebostatic axis was higher than the a-line at the external auditory meatus(mean 13.3 mmHg, standard deviation 6.1 mmHg, p-value <0.001, 95% CI 12.4-14.

2) The MAP measured by a-line leveled at the external auditory meatus was higher than that measured by cuff (mean 4.7 mmHg, standard deviation 10.8 mmHg, p-value <0.001, 95% CI 3.11-6.4) Pearson correlation coefficient for cuff and arterial line leveled at the phlebostatic axis was 0.68 with a p-value of <0.001, for a-line leveled at phlebostatic axis and at external auditory meatus was 0.9 with a p-value of <0.001, and between the a-line leveled at the external auditory meatus and the cuff was 0.66 with a p-value of <0.001

There were statistically significant differences in mean arterial pressure measurements by NIBP and a-line with the transducer leveled at the phlebostatic axis and the external auditory meatus. Since mean arterial pressure is used to calculate cerebral perfusion pressure, prospective studies looking at outcomes of patients by using two different techniques of mean arterial pressure measurement should be carried out.

The efficacy of administering single bolus doses of 14.6% or 23.4% hypertonic saline (HTS) to treat refractory intracranial hypertension has been demonstrated in the literature and has emerged as an important therapeutic option in treating these patients. However, many institutions lack experience with this therapy and there are few published studies evaluating the safety of repeated bolus dosing of HTS.

An observational retrospective review of patients admitted between January 2008 and July 2012 was conducted to evaluate the use of repeated dosing of HTS in patients with refractory intracranial hypertension. The primary objective was to evaluate the safety of repeated dosing of HTS assessed by documented adverse effects such as central pontine myelinolysis and fluctuations in serum sodium concentrations. Secondary objectives were to evaluate the efficacy of repeated dosing HTS in reducing intracranial pressure (ICP) and to compare the dose-response relationship of 14.6% and 23.4% doses.

Fifty-five patients were included for evaluation, each receiving an average of 8.9 (range 2-61) doses of HTS. A statistically significant increase in mean serum sodium concentration occurred with the administration of HTS (p<0.0001). No cases of central pontine myelinolysis were identified. The use of HTS was found to be effective based on decreases in ICP after administration (p<0.0001, mean ICP reduction: 10.1 mmHg, range 3-23.6 mmHg). The efficacy of 23.4% saline in decreasing ICP was not found to be significantly different than 14.6% saline (p=0.23).

Repeat bolus dosing of 14.6% or 23.4% HTS appears to be relatively safe and effective for treating refractory intracranial hypertension assuming there is frequent electrolyte monitoring and concomitant fluid management.

Intrathoracic Pressure Regulation (IPR) therapy non-invasively modulates pleural pressures to take advantage of the physiological benefits that occur by creating pressure differentials in the thorax. After each positive pressure breath, IPR lowers intrathoracic pressure to subatmospheric levels which enhances blood return to the heart and lowers intracranial pressure (ICP). We hypothesized that IPR therapy which has been previously shown to increase cerebral perfusion pressures (CerPP) and cerebral blood flow (CBF) in treated animals, would result in significantly improved CerPP and CBF when compared to control, untreated animals subjected to a focal brain injury.

In this study, 28 anesthetized animals were subjected to a focal brain injury by epidural insertion of an 8 French Foley catheter which was slowly filled with saline to simulate a traumatic brain injury with elevated ICP. Only animals that had a CerPP lower than 50 mmHg after balloon inflation were evaluated (n=23). Animals were prospectively randomized to no treatment or IPR therapy at a level of -12 cmH 2 O for four hours. ETCO 2 was maintained at 40 mmHg in all pigs. Results are reported as mean value ± standard error.

IPR therapy significantly increased CerPP (mmHg) throughout the study from 39.5 ± 1.7 to 43.7 ± 0.3 with IPR therapy but remained unchanged in controls (36.7 ± 1.4 to 35.4 ± 0.5), p<0.001. CBF (ml/100 gm/min), measured by Hemedex increased in the IPR group (34 ± 4 to 49 ± 7 @ 90 min) but not in controls (27 ± 1 to 25 ± 5 @ 90 min), p=0.01. Histology between groups was unremarkable. Cytokines IL-6 and INF-gamma levels were not significantly different between groups.

A significant increase in CerPP and CBF was observed during IPR therapy without any untoward lung effects. Additional animal and clinical studies are underway to fully evaluate the potential of this therapy to treat patients with compromised CerPP.

Financial Support: Drs. Metzger, Rees, Puertas are employed by the manufacturer of the technology studied in this work. Dr. Lurie is the founder and inventor of the company and technology.

Palliative care issues are often under-appreciated in the critical care setting. This is particularly evident in the presence of catastrophic loss, which is common in neurological critical care. Studies have shown that families who experience death of a loved one in the ICU setting, are at higher risk of post-traumatic stress syndrome. Therefore, through a quality improvement process, we developed a supportive care tool to enhance palliative care resources for patients and families in a 12-bed neurointensive care unit within a large tertiary care facility.

A supportive care tracking tool was developed based on a framework developed by the Center to Advanced Palliative Care (CAPC) I-PAL project. We created three categories (high/med/low) to stratify patient's level of supportive care needs. The categories were linked to specific clinical triggers that identified patients at high risk of death or disability. These clinical triggers were identified via daily assessments that were performed by advanced practice practitioners during rounds, and were linked to a specific set of supportive care tasks.

A total of 139 patients were assessed during a five month pilot interval. 35 fell into our 'high risk' category, 29 of which died in the ICU. There was a 52% increase in utilization of quality palliative care measures for those patients that died compared to baseline data. Overall unit length of hospital stay was reduced from 15.8 days to 13.3 days with reduction of ICU days from 8.6 to 6.2 days. Percentage of he key supportive care tasks completed was enhanced by 35% to 81%, with greatest improvement in care delivery in the high-risk category.

Utilization of an electronic supportive care rounding tool provided an effective way to identify patients at high risk of death or disability and to create an integrated process for the care team to meet palliative care needs of patients and families.

Clinical nurses providing end of life care identified inconsistent treatment plans when transitioning from aggressive medical management to comfort care in a Neurocritical care unit. Clinical nurses were surveyed; thirty percent of responders viewed provision of end of life care negatively. The reasons identified were ineffective team collaboration and communication, difficulty obtaining appropriate orders, lack of care standards, and lack of knowledge of symptom management. Lack of resources and support for families was another area of concern.

Methods A performance improvement project was established to standardize patient care, monitoring and documentation and palliative symptom management. A collaborative end of life care taskforce was established to standardize patient care, monitoring, documentation and palliative symptom management that can be seamlessly transitioned within the institution. A literature review identified best practices; a multi-disciplinary team collaborated to establish guidelines; and a postimplementation clinical nurse survey evaluated outcome. End of life care education was conducted to all clinical nurses and Neurocritical care team members.

The Neurocritical Care End of Life Order Set includes recommendations for pharmacologic symptom management, a terminal ventilator wean protocol, consults for family bereavement support, and patient care standards for assessment and documentation. Improved clinical nurse satisfaction with treatment plan consistency and symptom management has propelled this protocol organization-wide facilitating a smooth transition from ICU to hospice care.

A computerized order set facilitates patient safety and team communication through care transitions.

The majority of deaths following severe traumatic and non-traumatic brain injury may occur following withdrawal of life support. When early aggressive care is provided, decisions to withdraw care are often made after 1-3 months while clinical trial outcomes are sometimes assessed at 6 months. Our goal was to study the incidence of long-term functional recovery in patients with severe brain injury managed with tracheostomy. Need for tracheostomy may identify patients with particularly severe brain injury receiving early aggressive care.

We reviewed medical records of patients with traumatic brain injury (TBI), subarachnoid hemorrhage (SAH), intracerebral hemorrhage (ICH) and ischemic stroke (IS) with Glasgow Coma Scale (GCS)<9 admitted to the neuroICU who underwent tracheostomy. Additionally, diagnosis-specific thresholds identifying severity were used-Hunt-Hess>3 (SAH), ICH score>2 (ICH), NIH Stroke Scale (NIHSS)>15 (IS and ICH). Functional outcomes were estimated with modified rankin scale (mRS) and Glasgow Outcome Scale Extended (GOS-E) at 1-3 months, 6-12 months, 12-24 months and 24-36 months. Good outcome was dichotomized at mRS<3 and GOS-E>4.

A total of 109 patients with severe brain injury (SAH-39, ICH-29, TBI-23, IS-18) underwent tracheostomy mean 6+3 days from intubation. Adequate follow-up was available for 101, 81, 61 and 50 patients at 1-3m, 6-12m, 12-24m and 24-36m respectively. At the respective time periods, mRS<3 was seen in 5(5%), 31(38%), 25(41%) and 18(36%) and GOS-E>4 in 5(5%), 32(40%), 24(39%) and 18(36%). Of 16 patients with mRS>2 at 6-12m with adequate follow-up a further 8 (50%) improved to mRS<3 by 24m. Median time to decannulation was 33(IQR 21-53) days from tracheostomy.

While most patients (>90%) requiring tracheostomy are disabled 1-3 months following severe brain injury substantial proportions (38-40%) recover independence in activities of daily living by 6-12 months. Of note, up to half of patients dependent at 6-12 months may demonstrate delayed functional recovery by 24 months.

The inconsistent presence of diabetes insipidus (DI) and hypothermia (HT) in patients diagnosed as brain dead has fueled challenges to the whole brain formulation of brain death (BD). However, approaches to BD determination are highly variable and often incongruent with contemporary guidelines. Also, some co-morbidities and medical treatments can interrupt the expression of these complications. This may account for the inconsistent evidence for loss of hypothalamic function in BD patients. We sought to quantify the frequency of DI and HT at a medical center that regularly includes intracranial circulatory arrest (ICCA) for BD determination.

We reviewed the charts of all adult BD patients from 2007-2012 at our center. All aspects of the BD diagnosis, including brain imaging, clinical details, examinations, evidence of ICCA and any ancillary testing were reviewed. We then objectively determined whether these patients exhibited signs of DI and HT at or before the time of BD pronouncement with pre-determined criteria. 

1. DI and HT are present almost universally in BD patients diagnosed at our center, where ICCA affirmation is a priority. 2. These rates are higher than those previously reported and support the coherence of the whole brain formulation of brain death. 3. These findings also have implications regarding ideal approaches to BD determination.

The goal of multimodality monitoring is to provide real-time information regarding the relative health or distress of the brain and neurophysiologic decision support at the bedside to guide goal-directed therapy. In practice analysis of integrated multimodal physiologic data is either limited to what is available on a specific monitoring system or requires significant amounts of time and expertise if done in a batch analysis. We sought to demonstrate a flexible real-time application that could be generalized to any multimodality monitoring application.

Patient monitor and neuromonitoring data were acquired using Bedmaster EX. Mean arterial pressure (MAP) and intracranial pressure (ICP) were ingested every 10 seconds by IBM Streams, a streaming analytic middleware platform. For cerebral autoregulation assessment, the pressure reactivity index (PRx) and a locally-weighted regression line characterizing the relationship between MAP and ICP were continuously computed. Patient state changes were approximated by calculating financial support and resistance lines for each signal. A web interface to visualize the results was created using javascript and W3C technologies.

We have developed a demonstration real-time web application to enable clinicians to assess cerebral autoregulation status and identify potential patient state changes as they occur at the bedside. Most importantly, this infrastructure allows clinicians and quality control teams open access to raw data and the capacity to generate complex analytics using standard scientific development tools, including Matlab, R, or Python. Raw and processed data may be streamed simultaneously to bedside displays and long-term storage repositories.

We have demonstrated that real-time clinical analytics can be efficiently performed and displayed at the bedside. Future iterations have the potential to include signal processing of ICP waveforms or analysis of heart rate variability, and have the capacity to run predictive algorithms designed to identify secondary complications such as sepsis or delayed cerebral ischemia.

Financial Support: Dr. Schmidt received a faculty award from IBM Research and grant funding from TATRC. Dr. Schmidt and Dr. Mayer are Co-PI's of a grant award from the Dana Foundation.

Delirium is a common ICU complication and is associated with higher morbidity, mortality, and increased length of stay.

There is a paucity of data on nursing interventions to alleviate delirium. Sleep deprivation is implicated in the development of delirium. As such, enhancing sleep is a logical, readily feasible nursing intervention to prevent delirium. We report the development of an evidence based Sleep Care Bundle to promote sleep in a select group of ICU patients.

Literature review was completed by three bedside nurses in the Neuro-ICU and SICU at the University of Cincinnati Medical Center (UCMC) using the terms "delirium", "sleep deprivation", "intensive care units", and "intervention". Factors that contribute to sleep deprivation in the ICU setting and its effects on patient outcomes were identified. Subsequently, current sleep practices within our Neuro-ICU and SICU were reviewed. A Sleep Care Bundle was then developed, staff were educated regarding the bundle, and the effect of the bundle on patient care is concurrently being evaluated.

At the time of this writing, literature review and current practices of sleep care in the Neuro-ICU and SICU have been completed. Staff education has demonstrated an improved awareness of sleep deprivation and its effects on patients' wellness. Nursing interventions to improve patients sleep patterns in the ICU are provided within the Sleep Care Bundle guidelines. The impact of implementation of these guidelines will be assessed in the future.

Developing nurse driven directives and guidelines for sleep is feasible. A select group of stable ICU patients can be identified early and sleep promotion practices can be initiated.

Clevidipine is a third generation intravenous dihydropyridine calcium channel antagonist approved for treatment of acute hypertension when oral therapy is not feasible. Benefits include its tolerability, rapid onset and offset, and easy titration with predictable response. Clevidipine's formulation could be beneficial in fluid-restricted patients when compared to other antihypertensives, which deliver significantly more volume. Clevidipine use at our institution is limited to patients in selected intensive care units (ICUs). We present neurological ICU data evaluating effectiveness, safety, cost of clevidipine therapy, and appropriate use according to published guidelines.

Twenty patients receiving clevidipine between September 2011 and December 2012 were retrospectively identified. Indication for use, dose, duration of therapy, concomitant antihypertensive administration, blood pressure measurements, time to achieve blood pressure goal, and adverse events were assessed. Published guidelines were reviewed to determine whether clevidipine was an appropriate choice for the indication. Cost of therapy was calculated for each patient. Results were presented to the Neuroscience Subcommittee of the Pharmacy and Therapeutics (P&T) Committee.

Common indications for use included hypertensive crisis in patients with intracranial hemorrhage, acute ischemic stroke, brain aneurysms, and anoxic brain injury. Patients received clevidipine for a mean of 50.6 hours at a mean dose of 7.5 mg/hr resulting in a mean cost of $831.53 per patient. Patients reached their goal blood pressure in an average of 62.6 minutes. Fifteen percent of patients experienced hypotension, 77.8% of non-intubated patients experienced decreased oxygen saturation and 15% of patients had an increase in triglycerides. Based on current published guidelines, 35% of patients received clevidipine appropriately.

Based on this data, the Neuroscience Subcommittee of the P&T Committee recommended clevidipine as a first-line agent for patients with intracranial hemorrhage and unsecured aneurysms. However, due to its high cost, clevidipine's use is limited to 24 hours. Institutional guidelines are being developed.

Communication is an important safety and satisfaction indicator in the intensive care unit. Neurocritical care patients and families face unique communication barriers due to disorders of consciousness causing families to become surrogates for extended periods of time. Family meetings (FM) are a primary method of communication in Neurocritical care.

The Provider Family Meeting Attitude Survey (PFMAS) is a 10-item questionnaire focusing on family meeting attitudes and practices. The survey was administered to the Neurocritical Care Society (NCS) membership via their email list-serv in 2013.

Of 1460 

Training was uncommon among NCS practitioners who completed the survey. Many providers do not use structured formats which may enhance communication. Use of technological aids was low; video display, conference calling, Skypelike video conferencing, and computer-aided imagery may play an increasing role in FMs with the rise of technologicallydriven communication.

Many long-duration astronauts develop findings consistent with intracranial hypertension including headaches, papilledema, nasal congestion, and impaired taste and olfactory function. We prospectively evaluated olfactory function in patients with idiopathic intracranial hypertension (IIH) and the effect of six-degree head-down tilt (HDT) as ground-based space-flight analogs.: 

We demonstrated that IIH patients have profound impairment in olfactory detection and to a lesser extent in smell discrimination. There was a modest decrease in olfactory function during HDT compared to upright positioning for OT but not for smell discrimination.

Financial Support: This study was supported by a grant from the Center for Space Medicine, Baylor College of Medicine, Houston.

Currently, there are no reliable non-invasive methods for assessing intracranial pressure (ICP). Distortion product otoacoustic emissions (DPOAE) are evoked sound waves from the cochlea in response to specific externally delivered frequencies. Due to the connection of the cranial subarachnoid space and perilymphatic fluid via the cochlear aqueduct, changes in the ICP may affect DPOAEs.

We measured DPOAEs in 20 patients with mainly IIH or other neurological conditions simultaneously undergoing lumbar puncture. We measured DPOAE magnitude, phase angle shift, reflectance and impedence during the opening and closing ICP measurements respectively. We obtained Baylor College of Medicine IRB approval prior to this study.

Technical success was achieved in 90% (n=18) of patients. For data analysis, we divided subjects into three groups based on small, medium or large changes between opening and closing ICPs. For DPOAE magnitudes, we found significant increases (p<0.05) between pre and post-LP measurements respectively when comparing groups with large vs. medium, large vs small, or medium vs small ICP changes, mainly at DPOAE F2 frequencies between 700 to 1200 Hz. For the DPOAE phase angles, we found significant negative phase angle shifts of 0.1 to 0.25 cycles when comparing groups with large and medium ICP changes vs. those with small ICP changes, mainly at DPOAE F2 frequencies of 1000 to 2000 Hz. No significant changes in reflectance or impedence were seen.

We report for the first time a significant change in DPOAE magnitudes and phase angles immediately following CSF drainage. Specifically, DPOAE magnitudes increased, and phase angles significantly shifted in the negative direction. These findings are consistent with a previous study using DPOAE before and after body tilt as an analog for producing ICP changes. Future studies are warranted to develop correlation and ROC curves that relate DPOAE parameters with ICP changes, and identify the specific frequencies where ICP modulates.

The concept of brain death has evolved since first delineated in the 1960s [1] . The presence of established protocols internationally for brain death and their variations have not been evaluated recently, particularly since the more widespread practice of dead-donor transplantation.

An electronic survey was distributed to country-level physician representatives (January-June 2013) with reputed knowledge of neurocritical care via the NCS, published correspondence addresses, and personal contacts.

82 of 123 countries contacted responded (67%, with 28% from low-and middle-income countries) including neurointensivists (n=16, 19%), non-neurologist intensivists (n=18, 22%), neurologists (n=40, 49%), and physicians from other relevant subspecialties (n=8, 10%). Most reported a legal provision for brain death (n=58, 71%), with an additional 6 (7%) reporting an institutional protocol without a recognized legal provision. Countries with an organized transplant network were more likely to have a brain death provision compared with countries without (52/60 (87%) vs. 6/22 (27%), p<0.001). The three most common underlying diagnoses leading to brain death were reported to be traumatic brain injury (n=37, 58%), intracranial hemorrhage (n=14, 21%), and cardiac arrest (n=9,14%). Marked international variability was noted in requisite examination findings, the necessity and type of ancillary testing, and the minimum qualifications of physicians required for declaration. Distinct criteria for declaration in children were common (58%). Among respondents from hospitals without a provision for brain death, the most common reasons cited were 'no advanced technology/ICU care' and 'no physician expertise.'

This recent worldwide survey confirms that marked variation remains in the determination of brain death. We expand on the previously published literature [2] by adding data from 24 additional countries and exploring additional parameters. Further exploration of the reasons for this variability will be helpful in establishing consensus. [1] JAMA 1968; 205:337-340;  [2] Neurology 2002; 58:20-25.

This presentation will provide strategies for enabling professional engagement and active involvement of bedside nurses in championing key quality and outcome measures required for Comprehensive Stroke Center Certification. Practical advice regarding structure and processes will be discussed by the advanced practice nurse. Bedside nurses will share their experience of being empowered and professionally engaged. .

Multiple methods were employed including recruiting "stroke champions" from all units. This voluntary role includes attendance at stroke courses and performance improvement meetings, presenting case reviews, taking an active role in peer review, integrating patients and families into performance improvement and sharing this information at unit-level staff meetings These "stroke champions " also play an active role in reviewing our compliance rate for stroke process measures and offering suggestions for driving additional improvement.

The full stroke champion program was adopted in September 2012. We looked at two key nursing sensitive stroke care process measures and compared results for the 12-months prior to September 2012 and 7-months after deployment of the professional engagement strategy. Our compliance with stroke dysphagia swallow screening improved by 2% (from 84% to 86%); and compliance with discharge stroke education increased by 1% (from 93% to 94%). Additionally, comparing our patient satisfaction scores pre-and post-deployment demonstrate a 7% improvement in patient's perception of "instructions care at home" (from mean score of 85.0 to 91.1) and 5% increase in patient's perception of "extent felt ready discharge" (from mean score of 81.9 to 85.9). These two patient satisfaction measures are related to the discharge stroke education process measure

We believe the implementation of a "stroke champion" program is an effective strategy for reinforcing stroke education process measures and thus, improving stroke process measures and patient satisfaction scores.

Neuromuscular weakness causing ICU admission is most often due to inflammatory conditions such as myasthenic crisis or Guillain Barre Syndrome. Treatment is generally based on a clinical diagnosis, because immunological confirmation is delayed and electrophysiological testing may be limited. The diagnosis of rare myopathies is often made late or never, leading to inappropriate management.

Analysis of the diagnostic pathway in 3 cases of genetic myopathy manifesting acutely and diagnosed in the ICU using biopsy and genetic analysis.

A 58 year old male developed acute respiratory failure during a mountain hiking trip. He was diagnosed with a chest infection, failed to wean from the ventilator and was repatriated to the UK with suspected Critical Illness Myopathy. A history suggesting pre-existing diaphragmatic weakness was elicited; muscle biopsy and genetic testing confirmed Hereditary Myopathy with Early Respiratory Failure (HMERF) due to a Titin mutation. The patient weaned with nocturnal BiPAP. Case 2: A 69 year old female with 6 years of ophthalmoplegia was admitted for myasthenic crisis and underwent plasma exchanges, but symmetric ptosis, fixed weakness and slowly evolving dysphagia led to a muscle biopsy showing a vacuolar myopathy confirmed as Oculopharyngeal Muscular Dystrophy by PABPN1 mutation analysis. Case 3: A 75 year old man with a cardiac pacemaker and severe cardiomyopathy since 1 year and recent swallowing problems was admitted with severe sepsis. He had distal muscle wasting; both his sons had cardiomyopathy, so a muscle biopsy was obtained confirming myofibrillar myopathy.

Genetic myopathies are rare, but the sheer number of different conditions means that they are encountered more often than expected. These and other cases were all diagnosed in our unit during 1 year. Identifying clinical patterns suggestive of a rare neuromuscular disease can optimize use of EMG, muscle MRI, specialist immunohistochemistry and targeted mutation analyses in the ICU. 

Phenytoin is commonly used to treat or prevent seizures in critically ill patients. Due to its narrow therapeutic index, phenytoin drug concentration monitoring is required. This study evaluated the intrapatient utility of the initial total:free phenytoin concentration ratio to predict subsequent free phenytoin concentrations in critically ill neuroscience patients.

A retrospective electronic health record review was conducted by identifying patients who received phenytoin in the neuroscience ICU (NSICU) and had at least two sets of paired free and total phenytoin concentrations. The primary outcome was proportion of patient-specific total:free ratios within 20% of the initial ratio (constant). Secondary outcomes were to identify factors associated with non-constant phenytoin ratios and develop therapeutic drug monitoring recommendations.

100 patients with 556 paired total and free concentrations (mean total, free [SD]: 12.5 mcg/dL[5.3], 2.2mcg/dL[0.9]) were included. Overall, 193 (42%) post-initial pairs were non-constant occurring in 71 (71%) patients. Compared to patients with constant ratios throughout admission, non-constant patients were more often female and had more phenytoin concentrations drawn. The analysis did not show significance for any common co-medications affecting the phenytoin concentration ratios. Using patient-specific total:free ratios to predict patient-specific subsequent ratios resulted in poor correlation (R 2 =0.553). Evaluation of the Sheiner-Tozer equation in this population (mean albumin [SD] 2.7 g/dL [0.6]) showed moderate correlation (R 2 =0.758).

A majority of NSICU patients demonstrated non-constant total:free phenytoin concentration ratios. Individual total:free phenytoin ratios do not accurately predict subsequent free phenytoin concentrations. Using the Sheiner-Tozer equation to predict total PHT concentrations in the setting of hypoalbuminemia is an option in this patient population.

Tonic-clonic activity at onset (onset TCA) complicates 3% to 21% of cases of subarachnoid hemorrhage (SAH). The impact of onset TCA on in-hospital complications, including seizures, remains unclear. Regarding outcome, one study associated onset TCA with poor clinical outcome at 6 weeks after SAH, but no other studies have confirmed this relationship. This study aims to assess the impact of onset TCA on in-hospital complications, functional outcome, mortality, or epilepsy at 3 months.

Analysis of a prospective study cohort of 1479 SAH patients admitted to Columbia University Medical Center between 1996 and 2012. TCA within 6 hours of hemorrhage onset was identified based on accounts of emergency care providers or family witnesses.

TCA at onset was described in 170 patients (11%). Patients with onset TCA were younger (P=0.002), presented more often with poor clinical grade (55% vs. 26%, P<0.001), and had larger amounts of cisternal, intraventricular, and intracerebral blood than those without onset TCA (all, P<0.001). After adjusting for known confounders, onset TCA remained significantly associated with in-hospital seizures (OR 3.80, 95%-CI: 2.43-5.96, P<0.001), in-hospital pneumonia (OR 1.56, 95%-CI: 1.06-2.31, p=0.02), and delayed cerebral ischemia (OR 1.77, 95%-CI: 1.21-2.58, P=0.003). At 3 months, however, onset TCA was not associated with poor functional outcome, mortality or epilepsy after adjusting for age, admission clinical grade and cisternal blood volume.

Onset TCA is not rare, as it complicates 11% of cases of SAH. New and clinically relevant findings are the associations of onset TCA with in-hospital seizures, pneumonia and delayed cerebral ischemia. Despite the increased risk of in-hospital complications, onset TCA is not associated with disability, mortality and epilepsy at 3 months. The lack of association with outcome at 3 months is an important finding: despite the increased risk of in-hospital complications, onset TCA should not encourage withdrawal of care among SAH patients.

Seizure activity that persists despite acute administration of standard anticonvulsant therapy is termed refractory status epilepticus (RSE). RSE can be seen post-cardiac arrest and is a poor prognostic indicator. Therapeutic hypothermia (TH) has been shown to reduce neurological injury post-cardiac arrest, but is not routinely used in RSE from other etiologies.

Five consecutive patients with RSE admitted to our neurocritical care unit treated with TH (target temperature 33 °C) were retrospectively reviewed. Three patients had anoxic brain injury post-cardiac arrest and two had epilepsy secondary to remote traumatic brain injury. All five patients received similar medical treatment for status epilepticus.

The two post-traumatic epilepsy patients became seizure free after completion of TH and remained seizure free after rewarming. Both had good neurological recovery and were discharged for rehabilitation. Seizures persisted despite TH with the three post-cardiac arrest patients who later expired.

Patients with RSE caused by etiologies other than anoxic brain injury post-cardiac arrest may have better outcomes with TH, as supported by our two patients with RSE from post-traumatic epilepsy. The benefits of TH in RSE may be linked to the etiology of the seizures. Further studies are needed to evaluate RSE from a variety of causes to determine which groups would benefit from TH.

Super-Refractory Status Epilepticus (SRSE) is a Status Epilepticus that persists or recurs > 24 hours after start or withdrawal of anesthetics. Pentobarbital induced electroencephalogram (EEG) burst suppression is a common treatment. Reported duration vary from 7-244 hours with mortality rate of 13-77%.

A case series of 3 patients with SRSE treated with >96 hours of Pentobarbital.

A 41 year old male alcoholic with epilepsy presented with SRSE with sharp waves in the fronto-central region. He was treated with pentobarbital coma for 13 days and 4 antiepileptic medications (AEDs). He developed paralytic ileus, deep venous thrombosis (DVT) and pulmonary embolism (PE). He was discharged to home with gastrostomy tube and one AED. On follow-up, he was back to previous functional baseline. A 57 year old female with rheumatoid arthritis presented with multiple acute infarcts and on SRSE with right central parietal spike waves. Pentobarbital coma for 19 days with 5 AEDS was initiated. She developed DVT, PE, ARDS, paroxysmal atrial fibrillation, heparin induced thrombocytopenia and critical illness polyneuropathy. She was discharged to rehabilitation facility on 3 AEDs. On follow-up at 1 year, she was walking with a quad cane and back to functional baseline. A 23 year old male presented with Lepto-meningoencephalitis of unclear etiology and SRSE with high amplitude epileptiform spike and sharp wave complexes on the left hemisphere was placed on pentobarbital for 37 days with ketamine and 5 AEDs. He developed paralytic ileus, recurrent ventilator associated pneumonia, and provoked upper extremity DVT. He was transferred and reported to have been discharged to home to near normal functional baseline status.

It was shown that poor outcome is related to etiology of seizure and comorbidities. This series reflect good outcome regardless of seizure etiology. With meticulous medical treatment of complications and appropriate AEDs, prolonged pentobarbital coma is therapeutic and safe.

Only limited data describe the frequency, timing, or indications for endotracheal intubation (ETI) in patients with status epilepticus. A better understanding of the characteristics of patients with status epilepticus requiring airway interventions could guide clinical care. The objective of this study was to determine the characteristics of patients with prehospital status epilepticus (PSE) receiving ETI.

This study was a secondary analysis of the Rapid Anticonvulsant Medication Prior to Arrival Trial (RAMPART), a multicenter, randomized trial comparing intravenous lorazepam to intramuscular midazolam for PSE treatment. Subjects received ETI in the prehospital or ED setting at the discretion of caregivers. Demographic and clinical characteristics of intubated versus non-intubated subjects were compared using Student's t-test and chi-square or Fisher's exact test.

Of 1023 enrollments in the study, 218 (21%) received ETI. 204 (93.6%) of the ETIs were performed in the ED. ETI patients were older (52 vs. 41 years, p<0.001) and more likely to be male (60% vs. 53%, p=0.047). ETI patients were more likely to have ongoing seizures on ED arrival (48% vs. 28%, p<0.001) and to require rescue anti-seizure medication (26% vs. 18%, p=0.004). Hospital length of stay was longer for ETI (10 versus 4 days, p<0.001), and mortality was higher (7% vs. 0.4%, p<0.001). The most common reasons for ETI were respiratory depression (40%), depressed mental status (36%), and convulsions (16%). Most ETI (n=133, 61%) occurred early (prior to or within 30 min after ED arrival). Late ETI was associated with higher mortality (14% vs. 3%, p=0.002) than early ETI.

ETI is an important and common element of PSE care, especially among the elderly or those with refractory seizures. ETI often occurs early, and a minority of ETI is performed for respiratory depression. Indications for ETI in the PSE population warrant further investigation.

Background. Seizures refractory to third line therapy are also labeled superrefractory status epilepticus (SRSE). These seizures are extremely difficult to control and associated with poor outcome. Objective. To characterize efficacy and sideeffects of continuous infusions of pentobarbital (cIV-PTB) for SRSE treatment.

All adults with RSE treated with cIV-PTB between 5/1997 and 4/2010 at Columbia University were identified. We excluded patients post-anoxia and those receiving cIV-PTB for reasons other than RSE. We collected baseline information, continuous EEG findings, side-effects, and functional outcome at discharge and 1 year. Conclusions cIV-PTB effectively aborts SRSE and complications are infrequent but outcome in this highly refractory cohort of patients with devastating underlying etiologies remains poor. Phenobarbital may be particularly helpful when weaning cIVPTB.

There is widespread belief that beta-lactam antibiotics are associated with an increased risk of neurotoxic effects, with some attributing a higher association with cefepime due to its high penetration into the CNS. However there have been only limited case series describing this, and none using EEG findings as primary outcome. The purpose of this study is to more thoroughly investigate this hypothesis, by comparing the effects of cefepime and piperacillin/tazobactam, betalactams with comparable spectrums of use.

This is a retrospective case-control study of 175 ICU patients admitted to UPMC who received EEGs between March 2011 and November 2012. Exclusion criteria included patients with cardiac arrest and pre-existing refractory epilepsy. Patients were divided into three groups: those who received cefepime, those who received piperacillin/tazobactam, and those who did not receive any beta-lactams. (Beta-lactams must have been administered within 72 hours of EEG for those subjects to be considered.) The prevalence of neurotoxicity for each group was then determined, which was electrographically defined as the presence of seizures, PLEDs, focal sharp waves, TIRDA, or triphasic waves.

Of 175 patients, 13 were included in the cefepime group, 28 in the piperacillin/tazobactam group, and 91 in the control group. Prevalence of neurotoxicity was 38.4% with cefepime, 42.9% with piperacillin/tazobactam, and 19.8% in the control group. There was no difference between the cefepime and piperacillin/tazobactam group (p=1), while the difference between cefepime and the control group did not reach statistical significance (p= 0.1). However, the difference between the piperacillin/tazobactam and control groups did reach statistical significance (p=0.02).

Our results argue that the neurotoxic potential of cefepime may have been overstated, and that all beta-lactams may exhibit a similar degree of neurotoxicity. Further analysis will investigate the role of renal function and dose-dependence, as well as a subset focus on patients with primary neurological pathology.

Approximately a third of patients in neuroscience intensive care units (ICUs) experience subclinical seizures and, as a result, are at higher risk for poor outcomes. The use of continuous electroencephalography (cEEG) monitoring can help nurses detect seizure activity and initiate early prevention. Nurse competency in the use of cEEG is important to facilitate effective bedside monitoring. The objective of this study was to evaluate the effectiveness of a staff educational program aimed at improving the knowledge of nurses in the use of cEEG monitoring in adults.

A quasi-experimental pretest/posttest 1-group design was utilized. Neuroscience ICU registered nurses, whose experience ranged from 2 months to 24 years, participated in the study. Participants completed a pretest on seizure knowledge and the use of cEEG monitoring. Participants received a 4-hour educational session on the use of cEEG monitoring. Immediately after the program and again 1 month later, they completed a posttest. Test scores improved significantly from pretest to the first posttest (t = j15.093, p G .001).

Although there was a slight decline in the mean score from the posttest to the 1-month follow-up, posttest scores were significantly better than the pretest score (t = j12.42, df = 44, p G .001). Whereas years of experience correlated positively to the pretest score, after the intervention, no such correlation was evident. The results demonstrated that an educational program improved the competency of nurses in the use of cEEG with adult patients in a neuroscience ICU and that this knowledge was sustained over time.

Further research is needed to demonstrate the effectiveness of this intervention in other settings.

Generalized convulsive status epilepticus (GCSE) is associated with significant morbidity and mortality. The incidence of GCSE after elective admissions for neurosurgical procedures is unknown. The objective of this study is to determine the incidence and predictors of GCSE after elective neurosurgical procedures.

Elective hospitalizations for patients aged ≥18 years for all neurosurgical procedures (brain biopsy, incision and excision of brain and meninges, lobectomy, hemispherectomy, craniotomy, craniectomy, operations on globus pallidus and thalamus, ventriculostomy, ventricular shunt procedures and clipping) were identified in the Nationwide Inpatient Sample Database from 2002 to 2010 by using their respective ICD-9-CM codes. Patients with GCSE were identified using ICD-9-CM code 345.3. Logistic regression was used to determine the procedures associated with higher incidence of GCSE, reported in odds ratio with 95% confidence interval. We corrected for demographic information and baseline comorbidities (e.g. ischemic stroke, malignant tumors etc.) [1.08-3 .61] had a higher incidence of GCSE, and were included as covariates in the model. The procedures associated with a significantly high incidence of GCSE, after correction for the above covariates, were craniotomy and craniectomy (OR 1.64 [1.19-2.26 ], clipping (OR 2.66 [1.81-3.90 ] and lobectomy (OR 3.04 [1.74-5.33 ].

GCSE is a relatively uncommon complication among patients admitted for elective neurosurgical procedures. Craniotomy and craniectomy, clipping and lobectomy were associated with a higher risk for GCSE.

Status epilepticus (SE) is a life-threatening neurological emergency, and has been independently associated with poor outcome in cardiac arrest patients. Recent research demonstrates that the longer SE is left untreated, the greater the propensity for self-sustaining seizures and pharmaco-resistance. This raises the question of whether latency to initial treatment of SE (LTSE) is associated with better outcome in cardiac arrest patients.

This was a retrospective study from 1/1/2005-10/31/2012 of patients who underwent EEG at Hahnemann University Hospital with SE secondary to cardiac arrest. Baseline data included gender, age, code rhythm, code duration, LTSE, and APACHE score. The primary outcome measure was modified Rankin scale score (mRS). Secondary outcome measures included duration of SE, total number of antiepileptic drugs (AEDs) used for seizure cessation (tAEDs), days hospitalized, days in the ICU, and mortality. Statistical analysis was conducted using the Mann-Whitney-U-test for continuous variables, and the chi-squared-test for nominal data.

The cohort consisted of 46 cardiac arrest patients. Seventy-percent (n=32) underwent hypothermia, and 76% (n=35) demonstrated asystole. Age, gender, APACHE score and LTSE were similar between the hypothermia (+H) and nonhypothermia groups (-H), and for patients with and without asystole. For the -H group, LTSE≤10 and ≤30 minutes was significantly associated with shorter SE duration (p=0.006 and p=0.019), but not mRS, days hospitalized or in the ICU, tAEDs and mortality. There was no significant difference in outcome measures between asystole vs. non-asystole patients.

In patients with SE secondary to cardiac arrest who do not undergo hypothermia, LTSE ≤30 minutes is associated with shorter SE duration. This was not due to increased mortality, and suggests that shorter LTSE may modulate the pathophysiology of SE in these patients. However, poor outcome ultimately remains unchanged. This supports the contention that SE following cardiac arrest is independently associated with poor outcome.

Traumatic Brain Injury (TBI) is one of the leading cause of death and disabilities worldwide. Seizures are common complications after severe TBI and can precipitate secondary brain injury. Several studies detected high incidence of non convulsive seizures and non convulsive status epilepticus after TBI

We conducted a cross sectional study on 50 adult patients with severe TBI admitted to Alexandria main University Hospital. They were all under phenytoin therapy and underwent EEG 7days after admission to detect Non convulsive seizures. Then received 20 sessions of hyperbaric oxygen (5sessions per week) with pressure 1.5ATA.

statistically significant improvement in the GCS was detected along the first 2 weeks of therapy with the hyperbaric oxygen (the mean GCS in the first week was 6.7±1.1 improved during the second week to reach 9.1 ±1.5 ) associated with significant improvement in the EEG background. There were no statistically significant differences between the median rates of sharp waves discharge in the first EEG and the second one that was done 2 weeks after hyperbaric therapy (median rate of sharp waves in EEG1 was 30 and in EEG2 was 29, p = 0.161) also there was no statistically significant difference between rate of discharge of spikes and slow wave complexes between the first EEG and the second one done after starting hyperbaric oxygen therapy (median rate of spikes and wave in EEG1 were 30 and in EEG2 was 28, p= 0.524).

Hyperbaric oxygen therapy may have good effect on the EEG bachground and reactivity but has no effect on the control of non convulsive late post traumatic epilepsy Financial Support: None S228 Neurocrit Care (2013) 19:S1-S331 

Early MRI ischemia after subarachnoid hemorrhage (SAH) occurs in up to 40% of patients. Transient intracranial circulatory arrest at the time of aneurysm rupture, when ICP exceeds MAP and CBF is transiently low, may lead to watershed ischemia. We aimed to assess the frequency, predictors and impact on outcome of ictal MRI apparent diffusion coefficient (ADC) dark watershed infarcts.

In a prospective study, MRI (1.5T) was performed within 72 hours of SAH ictus (prior to clinical or radiographic vasospasm onset). DWI bright/ADC dark lesion locations were categorized as watershed (MCA-ACA, MCA-PCA or ICA), corpus callosal (representing deep penetrator watershed off the ACA and PCA) and non-watershed/callosal. Restricted diffusion due to hemorrhage or procedure was excluded. Demographic and clinical predictors of each location category were assessed using Chi-squared or Mann-Whitney U nonparametric testing. Logistic regression analysis was used to assess the association of early infarct pattern and 3-month outcomes.

Of 61 patients who underwent acute MRI, 276 DWI/ADC positive lesions were observed in 40 (66%) patients. Watershed pattern ischemia was seen in 31 (51%) of patients, 10 (16%) had callosal and 29 (48%) had non-watershed/callosal ischemia. Watershed and callosal infarcts accounted for 30% and 5%, respectively, of all infarcts. Poor Hunt-Hess grade, elevated BP at admission and poor APACHE2 score predicted watershed and callosal infarcts (all P<0.05). There was no association with loss of consciousness at ictus or global cerebral edema (known markers for intracranial circulatory arrest). Watershed infarct pattern was associated with worse 3-month cognitive status (telephone interview of cognitive status, OR 0.2, 95%CI 0.04-0.9, P=0.045), while other patterns were not.

Acute watershed and corpus callosum pattern ischemia occur in more that 50% of SAH patients and account for a third of all early infarcts. Watershed pattern ischemia is significantly associated with long term cognitive deficits.

Compared to surgical clipping, endovascular coiling of ruptured cerebral aneurysms is associated with better outcomes. The main objective is to evaluate the rates of hospital utilization as well as the outcomes of clipping and coiling treatments for ruptured cerebral aneurysm in teaching and non-teaching hospitals.

A large sample of treated ruptured aneurysms was extracted from the National Inpatient Sample (NIS) database using the ICD9-CM code for subarachnoid hemorrhage. Traumatic aneurysms, traumatic subarachnoid hemorrhage, arteriovenous malformation and dural fistulae were excluded. Treatment with surgical clipping or endovascular coiling was identified using the procedure codes. Hospital teaching status was linked to the NIS sample using the hospital identifier. Clip to coil ratio was measured for each hospital characteristic (hospital region, location, and bed-size) stratified by the hospital teaching status. Outcome analysis was comparative.

For the years 2000 through 2009, 19266 cases of subarachnoid hemorrhage were identified, 18287 included in the analysis. Sixty three percent of these aneurysms were clipped (n= 11509); and 37% (n=6778) were coiled. Teaching hospitals treated the majority of the aneurysms (85.8%) compared to non-teaching hospitals (14.2%). Among the nonteaching hospital, 76.1% underwent clipping compared to 60.8% in teaching hospitals. A composite clip to coil ratio in 8.36 in non-teaching hospital compared to 1.07 teaching hospitals (OR 2.0; 95% CI 1.86, 2.25; adjusted OR 2.0; 95% CI 1.79, 2.24). Post-operative stroke or hemorrhage was higher in the clipped population in teaching hospitals (P = 0.0002), while mortality rate was higher in the coiled population (0.0001). In non-teaching hospitals, the mortality and postoperative stroke or hemorrhage rates were similar.

Aneurysm treatment by clipping was more frequent in either hospital types. Clip to coil ratio was more balanced in teaching hospitals. Clipping-related complications were significantly present in the teaching hospitals along with outcome disparity.

Intrathecal Nicardipine (ITN) is potentially a useful intervention for the treatment of vasospasm after subarachnoid hemorrhage (SAH). We report our clinical experience with eleven patients who received ITN for the treatment of cerebral vasospasm.

This is a retrospective case-series. A clinical protocol for the use of ITN was established. SAH patients were considered for ITN therapy if patient had 1) Screening angiogram showing severe angiographic vasospasm 2) Symptoms of mild inattention and decreased alertness from vasospasm 3) Severe refractory vasospasm in conjunction with Intra arterial (IA) treatments. If patient had focal neurologic signs (cortical signs or motor weakness) ITN was only considered after angiography and IA treatment. ITN was discontinued when there was evidence of improvement of vasospasm (Computed Tomography (CT) Angiogram, Transcranial Doppler, clinical).

From November 2012 to May 2013 clinical information from 109 SAH patients were collected prospectively. Eleven patients were treated with ITN. All were Fisher group 3 with a median Hunt-Hess grade of 3. ITN was administered through an extraventricular catheter 4mg every 8h for an average of 3 days (range 1-6). ITN was started on median 8 th post bleed day (3-14). 10 patients had angiographic vasospasm and were treated with IA vasodilators prior to ITN. 8 patients received ITN based on presence of severe angiographic vasospasm. 3 patients received ITN based on clinical symptoms. ITN was well tolerated with no clinically significant increases in ICP, no hypotension and no ventriculitis. 3 patients needed to return to angiography for symptomatic vasospasm after treatment with ITN. 2 patients needed VP shunts. All patients survived until hospital discharge. We describe a demonstrative case of the usefulness of ITN in a 46 year old woman with severe vasospasm.

Intrathecal Nicardipine may be a safe and well tolerated treatment for cerebral vasospasm after SAH.

Delayed vasospasm after SAH is a common phenomenon that occurs in 2/3 of aneurysmal SAH patients. Vasospasm occurring immediately following SAH has not been well described in humans. We report on 11 patients who had aneurysmal rupture during catheter cerebral angiography, 10 of 11 of whom experienced hyperacute vasospasm -severe transient vasoconstriction immediately following aneurysmal rupture.

We retrospectively identified 11 patients who had aneurysmal rupture during catheter cerebral angiography between 1997-2009. Multiple vessel lumen diameters near and remote from the ruptured aneurysm were measured before and after rupture using either ImageJ software from captured images (JPEG), or using machine reference ruler from the angiography equipment. Measurements included an extradural reference vessel, allowing for calculation of a vascular diameter index (VDI), defined as the ratio between the intracranial vessel of interest and the reference vessel. Using either the VDI or absolute diameter, the change in vessel diameter was calculated (DVDI).

Significant reduction in the VDI was observed in 10/11 patients. For the 10 cases in which hyperacute vasospasm was observed, average difference in VDI after rupture was 0.09 (CI 0.04-0.13, p=0.001). The average change in VDI was -18.13%. Taking only the vessel segment that showed maximal narrowing in each case, average difference in VDI after rupture was 0.16 (CI 0.09-0.22, p=0.0003) and average change in VDI was -38.25%. Narrowing was maximal 39 minutes after rupture and began to resolve at about 1 hour. We observed delayed cerebral infarction in 55% of patients.

We report an under-recognized form of vasospasm occurring immediately after aneurysmal SAH in nearly all patients (10/11). The clinical significance of hyperacute vasospasm is uncertain, but may suggest a possible mechanism of early brain injury in the pathophysiology of vasospasm and consequently delayed ischemic injury. The presence of hyperacute vasospasm may suggest an earlier therapeutic target for pharmacologic intervention.

Outcome prediction after subarachnoid hemorrhage (SAH) at early time points is difficult. It is also still under discussion if the initial trama due to the SAH or later events such as vasospasm are finally responsible for the patient outcome. The objective of the present study was to evaluate if early S 100 CSF values might be a maker for the severity of the initial trauma and also an independent prognostic factor for patient outcome.

Patients with aneurysmatic SAH were included irrespective of endovascular or surgical aneurysm treatment if a CSF probe was sampled <48 h after sah. S 100 was immediately measured using an electrochemiluminescent assay.

Outcome was evaluated at hospital discharge and after 6 months using the Glascow Outcome Scale. Patients were divided in GOS 1-3 representing poor and GOS 4-5 representing good outcome. S100 levels were correlated to outcome and sensitivity and specificity were calculated. A possible correlation between S100 levels and vasospasm was examined using chi-square testing.

80 patients were included in the study, 32 patients had GOS values 4-5 and 48 GOS 1-3 at hospital discharge. S100 levels were clearly correlated to patient outcome, with a cutoff at 50 μg/l. Most patients with initial S100 levels below 50 μg/l were in the GOS 4-5, except 4 patients. Patients with S100 values >50 μg/l had poor outcome or died. Sensitivity was 0.86, specificity 0.73. Vasospasm was equally distributed in both patient groups showing no correlation between S100 values and vasospasm.

Our data clearly show a significant correlation between early S100 values and outcome after sah taking S100 > 50μg/l as cutoff for bad outcome with a sensitivity of 86% and specificity of 73 % . Since S100 was not correlated to vasospasm it seems to represent the importance of the initial trauma caused by the sah for patient outcome.

Autoregulation methods commonly employed are based on either inferential measures (eg, Prx, vasoreactivity measure) or are discontinuous. Mx, based on TCD, is difficult to do continuously. Brady et al demonstrated a NIRS StO2-based autoregulation index, COx. Our aim is to assess feasibility and demonstrate noninvasive continuous CBF based autoregulation assessment with a NIRS technique, diffuse correlation spectroscopy (DCS)(Yodh, etal).

DCS assessment of quantitative CBF with MAP changes was done over five two-hour sessions in a SAH patient and two two-hour sessions in a TBI patient. Changes in CBF were compared with eleven seconds earlier (from Budhoski etal) changes in arterial blood pressure, both compared to values at the beginning of the session, and a correlation coefficient was computed and called DCSx, over each two-hour monitoring session. The TBI patient was monitored on post injury days 15 and 16 and the SAH patient on post bleed days4,5,8,10 and 14.

The TBI patient underwent right hemicraniectomy with worse edema, contusions, and hypoperfusion on the right as seen on XeCTCBF and MRI. The SAH patient was HH3, Fisher 3. DCSx in the SAH patient over post bleed days 4,5,8,10 and 14 were, respectively, right: 0.33,0.72,0.52,0.28,0.22 and left: 0.46,0.41,0.72,0.66,0.81; and in the TBI patient over post injury days 15 and 16 right: 0.19, 0.01 and left 0.51, 0.61. These data suggest dysautoregulation in the SAH patient from days 4-14 and in the TBI patient on the left side, despite the hemicraniectomy on the right. Left-right heterogeneity was present in both cases.

We demonstrate an alternate approach to regional continuous CBF-based autoregulation assessment, using a noninvasive NIRS-based measure of quantitative CBF, in a TBI patient and in an SAH patient. Continuous CBF and associated continuous autoregulation assessment appears to be feasible using DCS. Regional DCSx assessment allows identification of heterogeneous autoregulation.

Financial Support: Dr Kofke was PI on NIH grant NINDS 1R21NS061857-01A2

Big endothelin-1 (bigET-1) is a substrate of endothelin converting enzyme-1 (ECE-1) and metalloproteinases (MMPs) and is the precursor to ET-1, a potent vasoconstrictor associated with vasospasm in SAH. We have previously found enzymatically active ECE-1 and MMP9 in CSF of SAH subjects and that elevated CSF MMP9 is associated with poor SAH outcome. We hypothesize that elevated CSF bigET-1 is associated with vasospasm and poor outcome in SAH.

We prospectively enrolled 49 consecutive SAH subjects with external ventricular drains placed for clinical indications, collected serial CSF samples, and evaluated outcome using modified Rankin scores (mRS) every 3 months. Poor functional outcome was defined as mRS>2. Angiographic vasospasm was defined as >50% reduction in caliber of any vessel on post-SAH day 7 cerebral angiogram. We compared CSF bigET-1 level and CSF ECE-1 activity on post-SAH days 1, 3, 5, and 7 by vasospasm and SAH outcome status. We used Bonferroni adjustment for multiple comparisons (significance threshold: p<0.0125).

Elevated CSF bigET-1 on post-SAH day 1 is associated with poor 3-month (p<0.0001) and poor 6-month SAH outcome (p=0.0120) after adjusting for age, Hunt and Hess (HH) and Fisher grades and aneurysm treatment modality by logistic regression. CSF bigET-1 showed strong positive correlation with CSF MMP9 (r=0.65, p=0.0086). There was a trend of association between elevated CSF bigET-1 on post-SAH day 7 and vasospasm (p=0.02) after adjusting for age, HH and Fisher grades and aneurysm treatment modality.

CSF bigET-1 elevation is independently associated with poor SAH outcome. Robust positive correlation between CSF bigET-1 and MMP9 levels suggests that a common process, such as leakage across the blood brain barrier, may have contributed to their release into CSF space. Future studies are necessary to understand the role of CSF bigET-1 in SAHrelated brain injury and to validate it as a potential biomarker for SAH outcome. 

Subarachnoid hemorrhage (SAH) is associated with significant changes in the fibrinolytic cascade implicated in delayed cerebral ischemia. Human data showed SAH-related brain ischemia is associated with polymorphisms in plasminogen activator inhibitor 1 (PAI-1) promoter region linked to higher plasma PAI-1 levels. We hypothesize that elevated plasma PAI-1 is associated with vasospasm and poor outcome in SAH.

We prospectively enrolled 61 consecutive SAH subjects, collected serial blood samples, and evaluated outcome using modified Rankin scores (mRS) every 3 months. We compared plasma PAI-1 levels by ELISA (R&D Systems) on post-SAH days 1, 5, 7 and 14 by angiographic vasospasm and by functional outcome status at 3-and 6-months. Angiographic vasospasm was defined as >50% reduction in vessel caliber on post-SAH day 6-8 angiography. Poor functional outcome was defined as mRS >2. Biomarker levels were compared using Wilcoxon rank sum test. Bonferroni correction was applied for multiple comparisons.

Subjects with (n=25) and without vasospasm (n=36) were comparable in gender, Fisher and Hunt and Hess (HH) grade distribution, aneurysm treatment modality, and incidence of hydrocephalus. Subjects with vasospasm had lower mean age (50 vs. 56). Elevated plasma PAI-1 on post-SAH day 1 was strongly association with vasospasm (3.28 vs. 1.82 ng/mL; p=0.0058). Plasma PAI-1 levels were not associated with functional outcome at 3 or 6 months. Elevated plasma PAI-1 on post-SAH day 1 remained significantly associated with vasospasm (p=0.0155) after adjustment for age and HH and Fisher grades by logistic regression.

Early elevation of plasma PAI-1 on post-SAH day 1 is strongly and independently associated with angiographic vasospasm in this small prospective SAH cohort and may be a candidate biomarker for vasospasm. Future mechanistic studies with simultaneous measurements of tissue plasminogen activator (tPA) and tPA-PAI complex and larger clinical studies are necessary to understand the utility of PAI-1 as potential biomarker in SAH.

Financial Support: This study was sponsored by the American Heart Association (10CRP2610341, Chou), Harvard CTSC/NIH (5KL2RR025757, Chou), NINDS (K23-NS073806 -Chou, R21-NS52498 -Ning, R01-NS48422 -Ning, R37-NS37074 -Lo, P01-NS55104 -Lo), and a departmental research discretionary fund (Chou).

Critically ill patients often exhibit a hyperdynamic response, which is linked with elevations in creatinine clearance (termed augmented renal clearance, ARC). Patients with subarachnoid hemorrhage (SAH) tend to be hyperdynamic, owing primarily to the physiologic response to SAH and frequent use of vasopressors and fluids for the treatment of cerebral vasospasm ("Triple H" therapy). No data exists investigating the presence of ARC in patients with SAH. We hypothesize that patients with SAH experience ARC, which significantly alters antimicrobial pharmacokinetics (PK), impacting the probability of achieving therapeutic drug concentrations of renally-eliminated antimicrobials. The purpose was to determine PK parameters of aminoglycosides and vancomycin in SAH patients at baseline and identify the impact of "Triple H" therapy.

We conducted a retrospective chart review of adult patients admitted with SAH between July 2008 and December 2012, who received aminoglycosides and/or vancomycin. PK parameters such as volume of distribution, half-life, and elimination rate were determined by utilizing plasma drug concentrations.

Eighty-seven patients met inclusion criteria. The mean age was 55.75 (±11) years and females accounted for 60.9% of the study population. The mean aminoglycoside dose (n=21) was 6.11 (±1.41)mg/kg, the volume of distribution was 0.51 (± 0.22)L/kg, and mean aminoglycoside AUC was 71.18 (±34.9)mg*L/hr. Durations of 8 hours or greater of undetectable aminoglycoside concentrations were common (62%). The mean vancomycin dose (n=72) was 15.96 (±2.54)mg/kg, mean trough concentration was 13.3(±7.94)mg/L, and the mean estimated half-life was 5.6 hours. No significant differences for either class of antimicrobials were evident in patients on "Triple H" therapy compared with those who were not.

SAH patients exhibit PK alterations related to increased clearance (possibly due to ARC) when compared to previously published data on normal subjects. The impact of "Triple H" therapy and other methods of optimizing cerebral blood flow requires further study.

Cognitive reserve hypothesis suggests that individual differences in variables like pre-morbid education, occupation, leisure activities, premorbid IQ influence how patients cope with brain pathology. Cognitive reserve hypothesis has been extensively validated in patients with Alzheimer's disease and to a lesser extent in patients with TBI and schizoaffective disorders. This abstract tests the cognitive reserve hypothesis in patients with SAH using education and occupation as proxy markers.

Patients were prospectively enrolled in a subarachnoid hemorrhage outcomes project (SHOP) between July 1996 to July 2012. Diagnosis of SAH was established by computerized tomographic scan or by a finding of xanthochromia on cerebrospinal fluid if the computerized tomographic scan is negative. We included patients with aneurysmal SAH who were ≥18 years. We excluded traumatic SAH and SAH from ruptured arteriovenous malformation or other secondary causes; admission more than 14 days after onset and a history of dementia.

We conducted a multivariate linear regression to examine the unique impacts of education and occupational attainment on cognitive outcomes in 555 patients. Predictors included age, sex, severity of SAH, presence of vasospasm and other relevant complications, length of hospital stay, performance on the telephonic interview for cognitive assessment scale (TICS) at discharge, education, and premorbid occupation. This set of predictors explained 65% of the variance in TICS improvement during the first 12 months following discharge (p<.001). Importantly, both education (B=0.45; SE=0.18; p=.01) and occupation (B=3.15; SE=1.25; p=.01) were significant, independent predictors of cognitive improvement.

Thus, it appears that both education and occupation represent separate aspects of cognitive reserve that might protect the brain from cognitive damage due to SAH. Trials designed to test interventions for improving cognitive outcomes after acute brain injury like SAH should probably control for differences in cognitive reserve markers to match patients appropriately for these trials.

Transfusion is a promising therapy for delayed cerebral ischemia (DCI) after subarachnoid hemorrhage (SAH). However, changes in red blood cell (RBC) deformability and oxygen-binding affinity, known to occur with prolonged storage, may attenuate the ability of transfusion to improve oxygen delivery (DO 2 ) and amount of oxygen offloaded (AV-DO 2 ) to vulnerable brain tissues. We evaluated whether fresh blood is better able to augment DO 2 and AV-DO 2 than blood stored for longer durations, especially in brain regions at high-risk for ischemia.

18 patients at-risk for DCI after SAH underwent 15 O-PET imaging before and after transfusion of one unit of RBCs. Five received fresh blood (stored ≤ 7 days) while 13 received older blood (stored ≥ 21 days), based on availability. Response between groups was compared using repeated measures linear models, globally and in brain regions with and without oligemia (low DO 2 and elevated OEF), correcting for baseline hemoglobin.

Rise in global DO 2 was non-significantly greater with fresh vs. older blood (4.9±1.2 to 5.6±1.2 ml/100g/min [+15%] vs. 5.8±1.9 to 5.9±1.6 [+2%], p=0.40) but this response differed regionally. Both fresh and older blood resulted in a significant 15-20% rise in DO 2 to the 13% of vulnerable brain regions with oligemia, but only fresh blood resulted in a rise in DO 2 to non-oligemic regions (p<0.001). While there was no difference in the rise in global AV-DO 2 with fresh vs. older blood (+20% vs. +9%, p=0.54), transfusion of older blood led to a significant drop in AV-DO 2 to regions with oligemia (p<0.001).

While both fresh and older stored blood augment oxygen delivery to vulnerable brain regions in patients with SAH, prolonged blood storage may have deleterious physiologic effects on oxygen offloading. How this impacts the clinical effectiveness of transfusion in cerebral ischemia requires further study.

Subarachnoid hemorrhage (SAH) patients may experience catecholamine release and activation of the renin-angiotensin system which could cause cardiac injury, myocardial infarction (MI), and death. We sought to determine whether preceding use of beta-blockers (BBs), statins, or angiotensin-converting-enzyme inhibitors (ACE-Is) or angiotensin II receptor blockers (ARBs) prior to SAH affected mortality; that is, whether such medications may be protective on overall survival similar to cardiology literature of patients with myocardial infarction.

We retrospectively reviewed the medical records of all non-traumatic (aneurysmal or other vascular) SAH patients admitted from March 2011 to December 2012 and reviewed the patients' admission medications for BBs, statins, ACE-Is and ARBs. Survival data was dichotomized as alive or dead, determined by chart follow-up. Statistical analysis was performed by a two-tailed Fisher's exact test comparing these medications against alive or dead status. Statistical significance was considered for P values <0.05.

We identified 134 SAH patients during the study period with a documented medication history. 13% (18/134) were on BBs prior to SAH; 22% (30/134) were on ACE-Is or ARBs prior to SAH; and 20.8% (28/106) were on statins prior to SAH. No statistically significant association was noted in mortality for patients with preceding use of BBs (OR 0.94, CI 0.28-3.10, P value =0.73), statins (OR =1.139, CI 0.432 to 3.002, P value =0.80) or ACE-Is or ARBs (OR 1.04, CI 0.3879 to 2.653, P value =1.0).

The data suggest that preceding use of beta-blockers, statins, or ACE-Is/ARBs is not associated with reduced mortality. The data was likely affected by smaller sample size of SAH patients on these medications. Larger studies may elucidate further if any statistically significant association exists.

The benefits of monitoring cerebral blood flow (CBF) in brain-injured patients are apparent. New techniques combining near infrared spectroscopy (NIRS) and indocyanine green (ICG) dye dilution to estimate CBF are available. However, transcutaneous NIRS with optodes applied over the skin is controversially discussed, because the signal is contaminated by contributions from extracerebral tissues (skin, skull, cerebrospinal fluid layer). Recently a new brain tissue probe for combined monitoring of intracranial pressure (ICP), CBF and oximetry with NIRS has been developed.

For NIRS 2 approaches are applied. 1. A conventional brain tissue probe for ICP monitoring is supplied with optical fibers (NeMo Probe®). Light is locally emitted into the brain tissue and a part, depending on the tissue properties, of it is received by detectors. 2. A plaster-based patch carrying optodes is attached to the skin (NeMo Patch®). Central venous injections of 0.2mg/kgbw ICG are performed. Measurement values are simulatenously collected with the NeMo Probe® and NeMo Patch® (NeMo System®, NeMoDevices, Zuerich, Switzerland).

12 measurements in parallel were performed. Mean blood flow values obtained with NeMo Probe (CBF Probe ) were 24.8 + 9.1 ml/100g/min with NeMo Patch (ECBF Patch ) 5.1 + 1.8 ml/100g/min. CBF values obtained with the NeMo Probe correlate significantly with blood flow values obtained with the NeMo Patch (CBF Probe vs. ECBF Patch r = 0.627 (p = 0.014); Spearman).

Blood flow values obtained with conventional NIRS correlate significantly with absolute CBF values obtained within the brain tissue. By simultaneous measurements with the NeMo Probe and NeMo Patch algorithms can be developed to quantify and subtract the contribution from extracerebral tissues from the signal obtained with transcranial NIRS.

Financial Support: Conflicts of interest:EK, JF and MM are stockholders in NeMoDevices.

We hypothesized that Body Mass Index (BMI) would impact the functional outcome in patients with subarachnoid hemorrhage (SAH).

We retrospectively reviewed all SAH cases between June 2008 and June 2012. We collected BMI and SAH severity scores by Fisher grade, World Federation of Neurosurgeons Scale (WFNS), Glasgow Coma Scale (GCS) at admission, and functional outcome by modified Rankin Scale (mRS) by death or discharge. BMI scores were categorized into the following ranges: <25.0, 25.0 -29.9, 30.0 -34.9, and ≥35.0.

In 274 SAH patients, the age ranged between 18-99 years (median 57), 61.7% females, and median BMI was 27. There was no association between BMI and GCS, WFNS, and mRS (all P≥ 0.12). mRS was worse for patients with a lower numeric BMI (Spearman's r: -0.13, 95% CI: -0.24 to-0.01, P=0.037), but this association weakened with categorical BMI. There was correlation between age and mRS (Spearman's r: 0.41, 95% CI: 0.30 to 0.50, P<0.001), WFNS grade (Spearman's r: 0.15, 95% CI: 0.04 to 0.27, P=0.011); and initial GCS (Spearman's r: -0.14, 95% CI: -0.25 to -0.02, P=0.023). The breakdown for age was >60 years for WFNS (P=.007) and GCS (P=.013) >65 years for mRS (P<.001).

BMI was not associated with severity of SAH measures and functional outcomes in patient with SAH. However, an older age was associated with worse functional outcome on mRS (>65 years), worse initial GCS, and WFNS (>60 years).

Investigate the incidence and correlation between the presence of neuron-specific enolase (NSE) in patients' blood and patient outcome after treatment of subarachnoid hemorrhage (SAH).

We retrospectively reviewed the medical records of all SAH patients admitted to our institution from June 2008 to June 2012. Bleeding severity on noncontrast CT scan at admission was assessed by the Fisher scale (grades 1-4), clinical severity by Glasgow Coma Scale (GCS) graded 3-15, and World Federation of Neurologic Surgeons Scale (WFNS) 0-5 at admission. Outcome was assessed by the modified Rankin Scale (mRS) 0-6 at discharge. We included patients who had NSE lab results obtained within 48hrs of admission, excluding traumatic SAH cases. The time from SAH onset or thunderclap headache was noted as well as the time the NSE was collected and reported. P-values for general tests of correlation resulted from Spearman's test of correlation. P-values were obtained using cutoff values for NSE ≤ 15 ng/mL versus NSE>15 groups.

Over a four-year period, we identified 309 nontraumatic SAH patients, 70 of which had NSE for analysis. 55% were female (39/70) with age ranging from 23-87 years old (mean 54). A significant higher proportion of patients with poor clinical grade at admission (WFNS 4-5, and GCS <7, p-value <0.001), and poor clinical outcome (mRS 4-6, p-value = 0.001) was found in patients with a higher NSE value cutoff value of ≥ 15 compared to ≤15 ng/mL. Fisher grade did not show a correlation to NSE levels or patient outcome (p-value = 0.69).

There is a significant association between NSE levels ≥ 15 ng/mL obtained in the blood within 48hrs after nontraumatic SAH and admission clinical grade and eventual patient outcome. The only measure of SAH severity that was not significantly associated with NSE level was Fisher grade (p≥0.43).

To characterize the patterns of hydrocephalus (HC) at the time of the external ventricular drain (EVD) clamp trial after aneurysmal subarachnoid hemorrhage (SAH) and subsequent need for ventriculoperitoneal (VP) shunt dependency.

We reviewed clinical and radiographic data of all SAH cases admitted to the Mayo Clinic Florida NeuroICU between June 2008 and April 2013. Cases were identified with symptomatic HC requiring EVD, and we documented Glasgow Coma Scale (GCS) (graded 3-15) at admission, 24 hours after EVD, and at discharge. Also those that underwent ventricular peritoneal shunting (VPS), length of hospital stay (LOS), modified Rankin scale (mRS) 0-6 at discharge, and intracranial pressure (ICP) during the EVD clamp trail. HC clamp trial failure was defined as ICP > 20mmHg as high pressure hydrocephalus (HPH), due to progressive obtundation or ventriculomegaly on post-EVD clamp trial head CT with an ICP > 5 but <20mmHg as normal pressure hydrocephalus (NPH) cases, clinical deterioration and progressive ventriculomegaly on CT 24hrs after EVD clamp trial with an ICP <5mmHg as low pressure hydrocephalus (LPH).

410 SAH patients were identified during the study. 74 (18%) cases of symptomatic HC required EVD placement. 54% females 46% males with a mean age of 56 and an age range of 27-88. 60 (81%) cases were NPH, 7 cases were LPH (9%), and 7 HPH cases (9%), and of those 53% (39/74) ultimately received VPS. Outcome (mRS) was statistically higher (p<0.0491) in HPH (6) cases compared to NPH (4), LPH (5.3).

SAH is associated with symptomatic HC in about 18% of cases, and half of those require VP shunting suggesting SAH blood products interrupt the normal cerebrospinal fluid absorption abilities. LPH occurs in a small number of cases yet, has much longer LOS perhaps due to delay in diagnosis (LPH 29, NPH 25, and HPH 16) (p=<.0025).

Headache (HA) and its treatments after spontaneous subarachnoid hemorrhage (SAH) are poorly characterized.

Retrospective review of non-traumatic Hunt and Hess (HH) grades I-III SAH patients admitted from 1/2011 to 3/2013. Demographics, clinical and radiographic features, medications, pain scores, and hospital course were recorded over days 0-14 post-bleed. Severity of HA was determined via the Numeric Rating Scale (0-10) and by total daily dose of analgesics.

Severe HA was defined as 2 or more days with maximum pain scores ≥8, or need for ≥3 different analgesics on 2 or more days. . Pain scores of 10/10 were reported by 33/77 (42.9%) patients after day 0 and by 7.8% on or after day 12. Medications for HA included acetaminophen (93.5%), dexamethasone (83.1%), oxycodone (71.4%), fentanyl (58.4%), morphine IV (55.8%), butalbital/acetaminophen/caffeine (40.3%), hydromorphone IV (33.8%), hydromorphone PO (33.8%), hydrocodone (10.4%), and ibuprofen (10.4%). Median pain scores peaked on day 0, trending slightly downward across the 15-day interval. Opiate analgesic usage peaked on days 4-5. 75.0% patients were discharged with opiate prescriptions.

Headache after spontaneous SAH had a non-linear association with severity (HH grade) and subarachnoid blood volume (Hijdra score), and was frequently severe, often necessitating multiple opiate and non-opiate analgesics. Many patients reported inadequate pain control as determined by persistently high pain scores; most required opiate discharge prescriptions. HA after SAH is a major problem requiring further study.

Introduction 6% Hydroxyethyl Starch 130/0.4 in 0.9% Sodium Chloride (Voluven®; 6% HES 130/0.4) is a colloid often used for fluid resuscitation in patients with subarachnoid hemorrhage (SAH), despite a lack of safety data for this use. The purpose of our study was to evaluate the effect of 6% HES 130/0.4 on major complications associated with SAH.

Medical records of all patients presenting between May 2010 and September 2012 with aneurysmal SAH were analyzed. Patients were divided in two groups based on the administration of 6% HES 130/0.4; HES group (n=57) and Non-HES group (n=72). The primary outcomes included a composite of 30-day mortality and three major complications associated with SAH [Delayed Cerebral Ischemia (DCI), Hydrocephalus (HCP) requiring cerebrospinal fluid (CSF) shunting, and rebleeding]. Secondary outcomes included median duration of Hospital and ICU stay.

The study groups were similar with respect to most characteristics. 

We observed a trend towards increased complications after SAH with 6% HES 130/0.4 administration. An adequately powered prospective randomized controlled trial into the safety of 6% HES 130/0.4 in this patient population is warranted to define the risks associated with HES administration.

Iron-dependent formation of reactive oxygen species has been implicated in the development of vasospasm and neuronal injury following experimental subarachnoid hemorrhage (SAH). Therefore we undertook this pilot study to explore the relationship between redox active iron (REDOX-Fe) and non-transferrin bound iron (NTBI) and the development of delayed cerebral ischemia (DCI) in patients with aneurysmal SAH.

Samples of cerebrospinal fluid (CSF) were obtained on days 1, 3, and 5. A fluorometric assay that relies on an oxidation sensitive probe was used to measure both NTBI and REDOX-Fe. We prospectively collected and recorded demographic, clinical, and radiological data, including the development of DCI (as defined by pre-specified criteria). One-way ANOVA, repeated measures ANOVA and the Wilcoxon Rank Sum test were used.

Median Hunt and Hess score on admission was 3.5 (range: 2-5) and median modified Fisher scale score was 4 (range: 2-4). Seven of 12 patients developed DCI (58%). Day 5 NTBI (7.88+1 vs. 3.58+ 0.8, p 0.02); peak NTBI (7.93±1.1 vs. 4.67±0.6, p 0.05); and mean NTBI (7.39+ 0.4 vs. 3.34+0.4 p 0.03) were significantly higher in patients who developed DCI. Trends towards higher mean REDOX-Fe and day 3 REDOX-Fe concentration (3.8 + 1 vs. 3.1 + 1 and 3.9 + 1.3 vs 3.1 + 1.3; p 0.08) in patients requiring with intra-arterial (IA) vasodilators and angioplasty were also found.

We found higher concentrations of NTBI in CSF of aSAH patients who subsequently developed DCI, suggesting a possible causal relationship between unbound iron and the development of vasospasm and DCI. These exploratory observations could have important therapeutic implications. Further studies are needed to validate these findings and to probe their clinical significance.

Financial Support: Internal grant. CV center, Cleveland Clinic

Intrathecal application of prolonged-release nimodipine microparticles (EG 1962, Edge Therapeutics Inc.) following experimental SAH in 2 different animal models demonstrated promising results. Therefore, we investigated the effect of EG-1962 in patients suffering from severe aSAH in a pilot series.

After obtaining informed consent from legal representatives, EG 1962 was administered in doses of either 40 (presumed non-therapeutic dose) or 100mg in patients suffering from high grade SAH (WFNS ° II-V and/or Fisher °III-IV). Administration was performed either intracisternally, following surgical aneurysm repair, or intraventricularly, via an external ventricular drain. Patients were closely monitored for occurrence of DCI using perfusion-CT imaging, transcranial doppler sonography and catheter angiography. In addition to common toxicity criteria, patients were specifically observed for increased occurrence of systemic hypotension. Additionally, the pharmacokinetic profile of CSF and plasma nimodipine levels was analyzed in all patients. Clinical outcome was assessed using the Glasgow Outcome Score (GOS).

To date, a total of 10 consecutive patients were treated. There were no adverse events, such as toxic drug effects or systemic hypotension. In patient #1 (treated with 40mg), the plasma-and CSF-levels were not measurable, as opposed to patients who received 100mg: Here, the plasma-and CSF-levels of nimodipine were within the therapeutic range over the analyzed period. Except patient #1, (treated with 40mg) all subsequently treated patients using the 100mg dose of EG 1962 demonstrated no clinical or radiological features of DCI or cerebral infarction on CT. After 6 weeks following SAH, GOS in patient #1 was 4, where as all other patients treated with 100mg and 200mg of EG 1962 reached a GOS of 5.

EG 1962 appears to be a promising new concept for prevention of DCI. Our encouraging initial results of the first clinical application of EG 1962 will be further investigated.

The benefits of monitoring cerebral blood flow (CBF) and oxygenation in patients with subarachnoid haemorrhage (SAH) are apparent. New techniques combining near infrared spectroscopy (NIRS) and indocyanine green (ICG) dye dilutions to estimate cerebral hemodynamics are available. The objective of this study is to develop a new brain tissue probe for combined monitoring of intracranial pressure (ICP), cerebral blood flow (CBF) and cerebral blood volume (CBV) with NIRS and ICG dye dilution.

For NIRS, conventional brain tissue probes for ICP monitoring are supplied with optical fibers (NeMo Probe®, NeMoDevices, Zuerich, Switzerland). Light is locally emitted into the brain tissue and a part, depending on the tissue properties, of it is received by detectors. Central venous injections of 0.2mg/kgbw ICG are performed. Regional values for the mean transit time of ICG (mttICG), CBF, CBV and brain tissue oxygen saturation (SbtiO 2 ) are calculated.

With a prototype of the probe measurements were performed in 7 patients. 52 measurement values could be calculated in 5 patients. No study related adverse events were observed. 24 pairs of repeated measurements under stable conditions of systemic hemodynamics and oxygenation could be performed. Mean values were for mttICG 5.74 + 0.27 sec, CBF 37.16 + 5.17 ml/100g/min and CBV 3.48 + 0.6 ml/100g. The coefficients of variation for repetitive measurements were 0.05 for mttICG, 0.16 for CBV and 0.14 for CBF, respectively.

NIR spectroscopy allows the simultaneous determination of multiple parameters with one single brain tissue probe. The new multi-parameter monitoring system can be applied as a powerful tool in detection and treatment of DCI after SAH.

Financial Support: Conflicts of interest:EK, JF and MM are a stockholders in NeMoDevices.

Continuous noninvasive measurement of CBF has not been previously reported. Thermodilution CBF has been reported but in a paucity of reports and without focus on variability. Jaeger, using thermodilution CBF in eight thirty-minute episodes(5 SAH, 3 TBI), in the process of comparing with P br O 2 , indicated a coefficient of variation of 0.44 suggesting that CBF is a rather dynamic variable. Using a noninvasive measure of CBF our goal was to describe the extent of variability in neuroICU patients with TBI and SAH.

Diffuse correlation spectroscopic optical monitoring was employed to evaluate quantitative changes in continuous bilateral CBF measurements in four two-hour monitoring sessions in two patients with TBI and 31 two-hour monitoring sessions in ten patients with SAH. Over each two hour epoch the mean change from baseline, standard deviation, coefficient of variation, range, and interquartile range was computed for left and right side measurements. Overall means of measures of variation for left and right sided data were computed. Data are reported as percent change from baseline at the beginning of each session.

The mean change in CBF from baseline was 13.7% with a mean standard deviation of 21.4% and coefficient of variation of 1.7. Mean IQR was 29% with mean range of CBF varying from 34% to 115% of baseline.

Many reports and clinical studies focus on solitary measures of CBF. Our data indicate that CBF changes dynamically in neuroICU patients to a potentially clinically significant extent such that single measurements, eg, daily TCD, CT/MR perfusion studies, or XeCTCBF, may give misleading information. The wide ranges of CBFs may reflect normal autoregulation associated with metabolic changes from nociception, anxiety, or sedation. Alternately it may indicate the effects of injury mediated dysautoregulation associated with physiologic changes in blood pressure.

Financial Support: Dr Kofke was PI on NIH research grant NINDS 1R21NS061857-01A2

Subarachnoid hemorrhage (SAH) from the rupture of an intracranial aneurysm shares key mechanical features with traumatic brain injury (TBI), including exposure to a sudden, global pressure wave generated by the arterial jet. It might, therefore, be anticipated that the diffuse mechanical injury of axons central to TBI pathophysiology is also an important component of cerebral injury after SAH. Elucidating these connections may lead to novel treatment approaches to both conditions.

We undertook a diffusion tensor imaging (DTI) study to understand the extent of axonal injury following SAH in a mouse model. We quantitatively compared changes in white matter anisotropy indicative of axonal integrity to histological evidence of axonal injury from the same tissue. We also examined the early time course of intracranial pressure (ICP) changes associated with SAH in this model system.

Our results show that ICP increases rapidly after SAH, peaking in approximately 2 seconds, and similar to that seen after various models of experimental TBI. DTI reveals a significant decrement in relative anisotropy in white matter regions close to the site of arterial rupture, with smaller reductions observed in more distant white matter structures. Histological analysis of amyloid precursor protein, known to accumulate in traumatically-injured axons, reveals multifocal axonal injury in a large halo surrounding the focus of subarachnoid bleeding. Silver staining also reveals protein accumulation in ipsilateral white matter structures as is seen following TBI.

These investigations reveal that axonal injury is a feature of early brain injury following SAH in this model system. The functional implications of this injury subtype, and its relevance to humans is the subject of future investigations. Further analysis of this phenomenon may help illuminate the processes underlying cerebral injury from SAH, provide new prognostic indicators, and suggest novel treatment modalities for this devastating condition.

Delayed cerebral ischemia (DCI) occurs in 20-30% of patients with ruptured aneurysmal subarachnoid hemorrhage (aSAH) and is a leading cause of death and disability. Modalities such as quantitative electroencephalogram (EEG) can be used to monitor vasospasm (Vespa, 1997; Claassen 2005) . The goal of this study is to determine if reduction in percentage alpha variability (PAV) is reduced before the angiographic detection of vasospasm or clinical signs of delayed cerebral ischemia.

166 consecutive patients from 2008-2012 with ruptured aSAH were monitored with cEEG at onset of admission to ICU. Patients who did not have cEEG or had vasospasm on their first angiogram were excluded. Vasospasm and DCI occurred in 30.1% of patients. Regional PAV's in the frontal, temporal and parietal lobes were evaluated from the time of admission to the end of ICU stay. Timing of the onset of reduction in PAV was compared with that of diagnosis of angiographic spasm and DCI.

cEEG was started at a mean of 20 (±8) hours after onset of SAH. Frontal lobe PAV had the most dynamic changes in PAV over time. PAV changes preceded angiogram-detected vasospasm with a mean lead time of less than 3 days. In addition, PAV remained up to 15 percentage points higher after intrathecal treatment for vasospasm, particularly after multiple treatments. Conclusions cEEG regional PAV in the frontal lobe is a helpful predictive indicator of DCI and angiographic detection of vasospasm.

Financial Support: None

Hydrocephalus after subarachnoid hemorrhage (SAH) frequently requires placement of a ventriculoperitoneal shunt (VPS). Little is known about the risk factors for delayed VPS placement in patients who initially had an external ventricular drain (EVD) successfully removed.

We performed a retrospective analysis of all SAH patients who had an EVD placed between 1/1/08 and 6/30/12 at our institution. We extracted demographic, imaging, and cerebrospinal fluid (CSF) data from medical records and analyzed risk factors associated with delayed VPS placement.

Of 93 patients who had their EVD removed, 12 (13%) required delayed VPS placement at a median of 54 days (interquartile range 15-75 days) after EVD removal--7 (58%) due to diagnosis of hydrocephalus on routine followup imaging, 4 (25%) due to clinical changes, and 1 (8%) due to leakage from the EVD site. After multivariate analysis, risk factors for delayed VPS placement included increased CSF protein within seven days of EVD placement (p=0.023) and increased third ventricular diameter prior to EVD removal (p=0.026).

Patients with increased protein at time of EVD placement and increased third ventricular diameter at time of EVD removal should be carefully monitored for development of late hydrocephalus.

There are multiple etiologies for failure while weaning an external ventricular drain (EVD) after subarachnoid hemorrhage (SAH), but there is little data on the relationship between etiology for wean failure and placement of ventriculoperitoneal shunt (VPS).

We performed a retrospective analysis of all SAH patients who had an EVD placed between 1/1/08 and 6/30/12 at our institution. For each wean step (defined as raising or clamping the EVD), we recorded success or failure (defined as lowering or opening the EVD) and etiology for failure (elevated intracranial pressure (ICP), clinical failure (due to headache or vomiting or altered mental status), leakage from the EVD site, or development of radiographic hydrocephalus), and we evaluated the relationship between etiology of wean failure and need for VPS.

Of 116 patients, 35 required VPS placement (30%). Patients who required VPS placement had a median of 2 (interquartile range (IQR) 1-4) wean failures and those who did not require VPS placement had a median of 1 (IQR 0-1) wean failures (p=0.001). There was no significant relationship between age, sex, Hunt Hess score, Fisher score, Glasgow Coma Scale, aneurysm location, aneurysm size, or aneurysm treatment method and need for VPS. There was no association between wean failure due to elevated ICP ever or failure due to EVD leakage ever and need for VPS, but there was a significant relationship between patients with at least one wean failure due to clinical changes and at least one wean failure due to radiographic hydrocephalus and need for VPS (p=0.007 and p=0.029, respectively). After multivariate analysis, there was only a significant relationship between wean failure due to clinical changes ever and need for VPS (p=0.018).

There is a significant association between wean failure due to clinical changes and requirement for VPS placement after SAH.

Basilar artery bifurcation aneurysm (BABA) is primarily repaired by endovascular approach. However, there are no clear strategies in treating complex and giant BABA, especially when presents with a poor Hunt & Hess (H&H) grades and intraventricular hemorrhage (IVH)

Retrospective review of cases from a prospectively maintained aneurysm database.

First patient was a 61 years old man who presented with H&H grade IV from a ruptured wide neck giant 33 mm aneurysm with IVH. The aneurysm was endovascularly secured. The IVH was treated with intraventricular tissue plasminogen activator (TPA). The clinical grade improved and underwent stent-assisted coiling of aneurysm. On 90 days patient became independent and asymptomatic (mRS 0), however follow-up angiograms in 6, 12, 18, 24 & 36 months demonstrated recurrences and underwent coiling of aneurysm. After the 5 th recurrences, patient underwent Y-Stent (enterprise) neck reconstruction and coiling of aneurysm. Follow-up angiogram demonstrated no recurrence and mRS remained 0. The second patient was a 48 years old man who presented with H&H II due to a giant aneurysm of 38 mm. Patient re-bleed and H&H beame V. Patient was immediately resuscitated. Repeat head scan demonstrated massive IVH without herniation. Emergent primary coiling of the aneurysm was performed. Patient received TPA through EVD resulted in complete resolution of IVH. In 36 hours the H&H reduced to II and pathent was extubated in 48 hours. 90 days mRS became 1; however, follow-up angiograms demonstrated recurrences and patient underwent Y-stent (enterprise) aneurysm neck reconstruction and coiling of aneurysm with obliteration of aneurysm. Patient remained independent and asymptomatic and on Plavix for stent protection.

Emergent endovascular securement of giant poor grade ruptured BABA aneurysm may open the door to treat the life threatening massive IVH using intraventricular TPA which may results in a good clinical outcome. Therefore, endovasacular repair of the poor grade ruptured BABA with IVH should be considered when poor grades are considered to be related to IVH.

Aneurysmal subarachnoid hemorrhage which accounts for about 30'000 cases per year in the US, has high morbidity and mortality. The incidence of unruptured cerebral aneurysms has been estimated to be much higher and up to 6% in general population. The pathogenesis of cerebral aneurysm formation has been discussed in the past but the exact molecular mechanisms are still to be elucidated. Alpha-1 antitrypsin, a potent protease inhibitor, has been implicated in pathogenesis of emphysema and liver cirrhosis, however; its role in aneurysm formation is debatable.

We measured serum levels of alpha-1 antitrypsin, alpha-1 antichemotrypsin and alpha-2 macroglobulin in 82 patients with cerebral aneurysms who were admitted to a single institution from 2006 to 2013. Genomic sequencing from those patients was also performed. Clinical characteristics and demographic data were obtained from our electronic medical record data base.

The serum alpha-1 antitrypsin level was less than 1mg/ml in 8 patients (9.7%) and less than 1.5 mg/ml in 32 patients (39%). The normal values were considered 1.5 to 3.5 mg/ml. The serum level of 1 to 1.5mg/ml was defined as low normal or borderline. A serum level of less than 1.5 mg/ml has been estimated to be seen in about 7% in normal population which is significantly lower than our aneurysm cases (39%).

A low serum level of alpha-1 antitrypsin is a potential risk factor for cerebral aneurysm formation. This emphasizes the theory of protease-antiprotease imbalance in pathogenesis of cerebral aneurysm formation.

The St. Michael's Hospital subarachnoid haemorrhage protocol was developed to reflect and expand on the latest evidence from the recently published Neurocritical Care Society's Multidisciplinary Consensus Conference, and Guidelines from the American Heart Association/American Stroke Association.

The protocol implementation consisted of: A) radiology technicians were in-serviced in the new CT Perfusion Vasospasm Protocol; B) A 1-hour lecture for neuro ICU nurses, preceded by a quiz including questions on practical aspects of aSAH management, as well as questions addressing nurses concerns. C) use of a standardized SAH admission packet. This intervention was carried out in a closed, 19-bed Trauma & Neurosurgical Intensive Care Unit at a Canadian academic hospital from February to June 2013. The intervention was evaluated using the first 20 patients admitted with the diagnosis of aSAH in 2012, as controls (chart audit).

Process measures: 47% of the audited charts contained documentation of a modified Fisher or WFNS score. After implementing of the revised SAH admission package, the frequency with which key information was recorded by the admitting physician, increased to 82% over a 4-month period. At the same time, the compliance with the new vasospasm monitoring protocol, using CT perfusion imaging reached 86%. Outcome measures:The mean time from aneurysm treatment to chemical VTE prophylaxis decreased from 7 to 2.5 days within 4 months. A trend to shorter ICU and hospital length of stay were found (unadjusted ICU stay from 13.6 to 12.6 and unadjusted hospital stay from 22.6 to 19.4).

The authors successfully implemented an evidence-based protocol for the management of patients suffering from aneurysmal subarachnoid haemorrhage, improving charting, DCI risk stratification and VTE prophylaxis. The implementation protocol data collection is an ongoing activity and relevant clinical data will be available shortly, specifically as it relates to clinical outcomes.

High quality care of aneurysmal subarachnoid haemorrhage (aSAH) is complex and demanding. Such care requires a multidisciplinary team spanning several medical specialities.The St. Michael's Hospital aSAH protocol was developed to reflect and expand on the latest evidence from the recently published Neurocritical Care Society's Multidisciplinary Consensus Conference, and Guidelines from the American Heart Association/American Stroke Association. This protocol is intended to aid in the daily dilemmas, doubts and uncertainties that are inherent when managing complex aSAH patients.

Clinical pathway flowcharts were developed to reflect the evidence behind the above mentioned guidelines coupled with our own group experience.

The first flowchart called admission, initial assessment and orders, was designed to mandate the implementation of hyperacute interventions aimed at preventing early rebleeding in unsecured aneurysms. The second flowchart, DELAYED CEREBRAL ISCHEMIA MONITORING PATHWAY, separates the patients into three different risk categories (designated: green, yellow and red), based on their clinical grade and Modified Fisher scores. The groups are monitored according to their risk category. Finally, the last flowchart is a new concept called CODE VASOSPAM, a timely approach to treat delayed cerebral ischemia, integrating and enforcing the team work of neuro ICU, neurosurgery and interventional neuroradiology.

The management of patient suffering from aSAH is complex, demanding and requires a multidisciplinary approach. This protocol combined the information of most recent guidelines, turning them into clinical practice pathways. It does not replace the physician's judgment in individual cases, but may be considered reasonable and current approaches to the management of this complex critically ill population. It is intended to foster a coordinated, integrated approach among the multidisciplinary team caring for aSAH patients, with the ultimate goal of improving long-term neurological outcome.

Subarachnoid hemorrhage (SAH) patients are at high risk for venous thromboembolism (VTE). Given the paucity of literature, the emphasis on chemoprophylaxis varies across centers. Our aim is to describe the clinical outcomes of VTE following SAH.

We identified patients with SAH and VTE in the Nationwide Inpatient Sample Database from 2007 to 2010 using ICD-9 codes. We compared patient demographics, hospital characteristics and outcomes between the VTE and non-VTE groups. The primary outcomes were in-hospital mortality and home discharge. Secondary outcomes were length of stay (LOS) and total hospital charges. Pearson's Chi-square test and Wilcoxon-Mann Whitney tests were used for categorical and continuous variables respectively. We also built multivariate logistic regression models to assess VTE outcomes after adjusting for confounders. 

VTE is associated with worse outcomes in SAH patients, with lower home discharge rate, longer length of stay and higher hospital costs. SAH patients with VTE had more co-morbidities. However, VTE was not associated with increased mortality rates. Lack of availability of post discharge follow-up and admission SAH severity grades are limiting factors. Prospective confirmation is warranted.

We sought to understand how referral bias affects the rate of EEG seizure detection following aneurysmal subarachnoid hemorrhage (aSAH). In aSAH patients of high clinical or radiographic grade, we evaluated EEG seizure detection rates before and after institution of an aSAH EEG ischemia monitoring protocol in which EEG is recorded independent of clinical suspicion for seizure.

We retrospectively reviewed electronic medical records and clinical EEG for 54 aSAH patients with non-traumatic SAH presenting either with Hunt Hess (HH) 4-5 clinical grade or Fisher (F) 3 radiographic group. Univariate analysis of variables influencing rates of electrographic seizure detection included age, aneurysm location, HH/F grade, and presence or absence of clinical suspicion for seizure, intraparenchymal hemorrhage, intraventricular hemorrhage and hydrocephalus. A univariate threshold of p <.15 was selected for inclusion in a multivariate logistic regression model for predicting seizure occurrence.

Compared to a 23.1% (n = 6) rate of electrographic seizures in 26 patients referred for EEG due to a clinical suspicion for seizure, electrographic seizures were less frequent among those 28 referred without clinical suspicion (7.1%; n = 2) for seizures. Among all 54 high-grade aSAH patients, irrespective of EEG referral, 14.1% (n=8) developed seizures. MCA aneurysm location and clinical suspicion for seizure qualified for entrance into the multivariate model (P <.01), but only age remained significant for predicting seizure after logistic regression was performed (OR = 1.10, 95% CI -0.19 to -0.20; p .03).

Even when seizures are clinically unsuspected, EEG in aSAH patients of high clinical or radiographic grade detects a modest and non-trivial rate of seizures. While referral bias may enrich rates of seizure detection, expanding EEG monitoring to include all high-grade patients may diagnose electrographic seizures that may otherwise remain undiscovered. Presence of seizures is nevertheless associated with poor outcome.

Fever after spontaneous subarachnoid hemorrhage (SAH) has been associated with worse clinical outcomes. The importance of timing of fever has not been well characterized.

A prospective cohort study was conducted in SAH patients with body temperature recorded for first 6 days of hospitalization. Baseline demographics, Hunt-Hess grade, Fisher grade, occurrence of delayed cerebral ischemia (DCI) and modified ranking scale (mRS) at discharge were adjudicated. Daily maximum temperature (Tmax) and a fever cut off of 38°C and 38.3°C were used. Univariate and multivariate logistic regression analysis were used to determine the ability of different fever cut offs and variables to predict clinical outcome.

A total of 98 patients were included in our analysis. 57% and 40% of patients had fever using the cutoff values of 38°C and 38.3°C, respectively. Findings were similar using either cut off. Fever was associated with DCI, worse mRS at discharge, and longer hospital stay. When controlled for baseline demographics and Hunt-Hess or Fisher group, fever (38°C) was a significant predictor of DCI (OR-1.7; p=0.02), mRS (OR-1.7;p<0.01) and longer hospital stay (OR-1.7;p<0.01). Tmax of patients with DCI was found to be significantly higher than those without DCI (p=0.04). Tmax between DCI and no-DCI groups separated on bleed day 2. In multivariate analysis Tmax on bleed day 2 and change in Tmax between bleed days 1 and 2 predicted occurrence of DCI, even after adjusting for covariates. No difference in WBC counts was seen between DCI and no-DCI groups or between patients with good versus poor outcomes. In general, there was a downward trend in WBC count in the first 6 days after SAH.

Early low grade fever predicts DCI and worse clinical outcomes. This suggests that inflammation after SAH occurs early in the time course of disease, before the maximal vasospasm period.

Nimodipine is used to decrease delayed ischemic neurologic deficits in patients who have suffered aneurysmal subarachnoid hemorrhage (aSAH). Nimodipine has not consistently decreased angiographic vasospasms; however, it has improved neurologic outcomes after aSAH. The FDA approved dosing is 60 mg every 4 hours for 21 days. Shortened courses of nimodipine have not been fully evaluated in patients with all severities of aSAH. The hypothesis of this study is that a shortened course of nimodipine is not inferior to the 21 day course of therapy that has shown benefit in historical trials.

A retrospective, single center electronic health record review including patients admitted to the University of Cincinnati Medical Center (UCMC) with aSAH and received at least one dose of nimodipine. Patients at UCMC generally receive a shortened course of nimodipine therapy, up to 14 days. The patients that received shortened courses of nimodipine at UCMC will be compared to a historical cohort. The primary outcome is cerebral infarction. Secondary endpoints include compliance with nimodipine therapy, rates of angiographic vasospasm and Glasgow Outcome Scale.

The UCMC cohort contained an older population with higher World Federation of Neurological Surgeons (WFNS) aSAH scores than the historical cohort. The rates of cerebral infarction on computed tomography were higher in the UCMC cohort when compared to historical controls (36% v 22%; p<0.011). The rates of vasospasm were similar between groups (21% v 19%; p= 0.846). Patients admitted to UCMC received a mean+SD of 10.3+3.2 days of nimodipine and averaged 64.2% of the daily FDA recommended dose of nimodipine. The historical cohort had significantly more good recoveries than the UCMC cohort (72% v 47%; p<0.001). No other statistical differences were found for other endpoints.

An increase in cerebral infarctions was shown in the UCMC cohort possibly due to increased age and WFNS score.

The presence of coagulopathy after subarachnoid hemorrhage has been suggested by prior studies demonstrating activation of both the coagulation and fibrinolytic systems. We sought to determine whether there is evidence of hypercoagulability defined by thromboelastography (TEG), a point-of-care test that allows for rapid global assessment of coagulation. TEG analyzes whole blood, not plasma, which better accounts for the effects of cellular components on hemostasis.

Ten patients with moderate-to-severe SAH (Fisher group 3 and >Hunt and Hess Grade 2) were prospectively enrolled in an IRB-approved observational study of serial TEG. TEG analysis, using kaolin activated citrated samples, was performed on post-bleed days (PBD) 1, 3, 5, 7 and 10 and compared to 15 sex matched control subjects. Mann Whitney nonparametric tests were performed to determine whether indices of hypercoagulability [maximum amplitude (MA), alpha angle (αA), total thrombin generation (TTG)], were significantly different from controls.

Mean age of patients was 59.1+/-11.6 years. The MA was elevated on PBD 5 [71. 1 (5.8) 

TEG identifies a transient hypercoagulable state that peaks around post-bleed day 5-7 in patients with SAH. This hypercoagulable state is primarily due to increased platelet activation, although there may be contribution from enzymatic hypercoagulability in the latter part of the hypercoagulable window.

Aneurysmal subararachnoid hemorrhage (aSAH) patients are monitored for 10-14 days. When delayed cerebral ischemia in the form of vasospasm (VSP){Vergouwen2010} is detected, it is treated with hypertensive (and sometimes endovascular) therapy{ConnollyJr2012}. Current recommended classification methods are resource-intensive, relying on specialty-trained professionals (nursing exams, TCDs, perfusion imaging){ConnollyJr2012}. Routinely monitored variables such as cerebrospinal fluid drainage volumes (CSF), sodium and glucose, intracranial pressures (ICP), blood pressures (BP), and heart rates (HR) are not used to predict VSP. We hypothesize that these passively-obtained features yield as much information as resource-intensive features to classify VSP.

We studied 81 Fisher grade >2 patients with secured aneurysms presenting within post-bleed day (PBD) 0-1 without early VSP. VSP Class was defined as angiographic VSP warranting endovascular treatment (yes/no). Resource-intensive variables (TCD/exams) and passively-obtained variables (CSF/BP/HR/ICP/glucose/sodium) from PBD0-2 were abstracted from hospital records. Naïve bayes (NB) and logistic regression classifiers were trained on selected feature sets of these variables. 10-fold cross validation prevented over-fitting. Performance of trained classifiers was evaluated using area under the receiver operator characteristic curve (AUC) and F-measure. Ablation analysis determined incremental utility of each variable.

.2% developed VSP. Some isolated features of passively-obtained variables produced classifiers that could predict VSP better than resource-intensive measures (e.g. AUC and F-measure of NB:glucose > NB:TCDs). Final ablation analysis considered larger feature subsets: NB classifier trained on passively-obtained features achieved AUC 0.708 and F-measure 0.636, outperforming NB classifier trained on resource-intensive features which achieved AUC 0.501 and Fmeasure 0.349.

Data-driven analysis of passively-obtained clinical data seems to predict VSP at least as well as current targeted monitoring techniques after aSAH. Our methods are not enabled by hospital information technology infrastructures. We are validating our model and developing a modular decision support tool, with plans to compare it against prospectively recorded cognitive assessments and clinical decision making.

Financial Support: This research was supported in part by NSF CNS-1035715 and NIH 1U01EB012470-01.

Hypoalbuminemia has been reported in critically ill patients including aneurysmal subarachnoid hemorrhage (aSAH). Existing data on hypoalbuminemia and cerebral vasospasm incidence is conflicting. Hypoalbuminemia is identified as marker for increased morbidity and mortality in ICU patients, but there are no data examining acute change (loss) in serum albumin as a function of cerebral vasospasm and 6-month functional outcomes in aSAH.

Retrospective chart review from July 2010 to June 2011. Subjects were included if they were adults admitted by postbleed day 2, with at least two serum albumin levels (admission and follow-up). Acute albumin loss was defined as >1.0g/dl drop in serum albumin compared to the baseline (admission albumin). Functional outcome was assessed by modified Rankin Scale (mRS) scores at 6-months. Analyses were completed using SAS v9.3.

Seventy-five patients were admitted with aSAH during the study period. Of the 31 who met full inclusion criteria, 18 (58%) were seniors (over 65 years old). Mean baseline serum albumin was 4.1 g/dl (IQR = 3.8 to 4.4), mean follow-up serum was 3.2 g/dl (IQR=2.9 to 3.4). Acute albumin loss was observed in 12 (38.7%) patients. Clinical vasospasm was observed in 6/31 (19.35%) patients. There was no association between being seniors and the odds of experience acute albumin loss (X 2 = 0.6; p=0.44). For patients with acute albumin loss, there was no increase in the odds of experiencing vasospasm (X 2 =0.09; p=0.76), nor of a bad (mRS>2) functional outcome (X 2 = 0.2; p=0.89).

Acute albumin loss as defined in the present study was not associated with adverse short-term or long-term (6-month) clinical outcome in aSAH patients. The results may reflect bias from a small sample size and should be confirmed in a larger cohort.

Systemic inflammation elevations and resulting reductions in heart rate variability may be associated with infectious complications and delayed cerebral ischemia (DCI) from cerebral vasospasm after subarachnoid hemorrhage. We sought to determine if monitoring heart rate variability (HRV) would facilitate early detection of these complications 24 hours in advance.

Continuous EKGs recorded from 292 of 449 consecutively admitted SAH patients between 2006 and 2011 were analyzed using standard methods to quantify HRV. Clinical variables and HRV metrics were time-locked to the ICU day prior to the development of the first secondary complication (i.e., sepsis, pneumonia, urinary tract infection, DCI) or post bleed day six. All missing values were globally replaced, nominal attributes were transformed into binary ones, and all attributes were normalized. A bagging meta learning scheme applied to cost sensitive random forest classifiers was used on this data using a stratified 10-fold cross-validation approach to estimate the error rate.

Results suggest that patient that will have an infectious complication or DCI can be identified up to 24 hours in advance with high sensitivity and specificity (Area under the curve = 0.76; True Positive Rate = 0.70; False Positive Rate = 0.33; Precision = 0.70; Recall = 0.70, Accuracy = 69.5%). Important predictor variables include admission Hunt Hess grade, modified fisher score, cerebral edema on admission CT, TCD, SIRS, and heart rate variability (e.g., RMSSD and sample entropy).

Heart rate variability monitoring may be a viable method to help identify patients in the preclinical stage of infectious complications and delayed cerebral ischemia. A validation study confirming these findings is needed before it could be tested as an early warning monitor in a clinical setting.

Financial Support: Dr. Schmidt received financial support through the Columbia University Clinical Translational Science Award KL2 program and also received a faculty award from IBM Research.

We previously reported that cranial accelerometry can detect a cranial "bruit" in patients with cerebral vasospasm. This is a shift to higher frequencies of the vibratory signature of pulsatile blood flow through the skull. We sought to test the accuracy of this technique to detect cerebral vasospasm in a prospective blinded study.

Skull accelerometry was performed using a prototype device designed by Jan Medial, Inc. (Mountain View CA). Paired TCD recordings and accelerometry epochs (typically 15 minutes of recording) were obtained in consecutive patients with subarachnoid hemorrhage undergoing TCD recordings for surveillance of vasospasm in whom informed consent was obtained. Accelerometry signals were converted to the frequency domain and the energy at frequencies above 100 Hz was quantified as a measure of vasospasm presence. TCD defined spasm was dichotomized as moderate/severe, or mild/absent within each insonated vascular segment.

A total of 130 accelerometry recordings were obtained in 38 subjects with subarachnoid hemorrhage who had paired TCD recordings. 5 of these recordings were classified as moderate/severe within a vascular segment. Accelerometry measurements identified 4 of these 5 (sensitivity 80%) and reported 20 false positives (specificity 39%). This model will be tested on 50 prospectively obtained studies already collected and the final accuracy of this technique will be reported.

Highly sensitive skull accelerometry can detect cerebral vasospasm with high sensitivity and validated results on a large cohort of patients will be presented. This non-invasive methodology holds promise as an alternative to TCD for detecting cerebral vasospasm caused by subarachnoid hemorrhage.

Financial Support: WSS has received a grant from Jan Medical, Inc.

Symptomatic vasospasm is a common complication affecting morbidity and mortality after subarachnoid hemorrhage (SAH). We sought to identify predictors of treatment failure among SAH patients treated with hypertensive hypervolemic therapy (HHT) for symptomatic vasospasm after SAH.

We retrospectively analyzed all SAH patients with symptomatic vasospasm among 1566 patients enrolled in the Columbia University SAH Outcomes Projects between July 1996 and August 2012. Treatment response was adjudicated in weekly meetings of the study team based on serial clinical examination. Treatment failure was defined as lack of any clinical improvement after volume leading and induced hypertension

Of 230 symptomatic vasospasm patients, clinical response was recorded in 198. Mean age was 53 years, 76% were female, and 50% were white. After volume loading and induced hypertension, 29% experienced complete resolution of their clinical deficit, 50% experienced partial improvement, 16% experienced no improvement, and 5% were worse. One hundred eighty-seven patients received follow up angiogram. Failure of response to medical therapy was associated with changes in mental status (68% vs 50%, P=0.037) and angiographic evidence of more than 1 arterial segment involvement (92% vs 43%, P<0.001). There were no significant differences between the groups with regard to Hunt-Hess grade, modified Fisher Scale, timing of onset of symptoms, or baseline vital signs.

Symptomatic vasospasm patients presenting with changes in mental status and multi-vessel syndromes are at high risk for medical treatment failure. Urgent referral for endovascular therapy should be strongly considered in these patients.

Targeted endovascular treatment (intra-arterial vasodilators and/or angioplasty) of symptomatic cerebral vasospasm and cerebral ischemia associated with aneurysmal subarachnoid hemorrhage remains with worrisome side-effects and controversial efficacy. Milrinone, a selective phosphodiesterase III inhibitor has shown limited but promising results as an effective cerebral vasodilator with inotropic properties.

A review of all patients with symptomatic and angiographically-proven vasospasm who received intra-arterial milrinone between January 2012 and January 2013 at a single neurological center was undertaken. Quantitative grading of angiographic vasospasm (mild, moderate and severe), evaluation of angiographic response following targeted treatment and any possible complications were assessed by a blinded reviewer. Immediate clinical neurological response was noted and 6 -12 month follow-up using modified Rankin scale was performed.

Five patients (3 males and 2 females) with a mean age of 53 years were identified. On initial presentation of symptomatic vasospasm, all patients were treated with high dose continuous intravenous infusion of milrinone, blood pressure augmentation and maintenance of euvolemia. Due to recurrence of symptomatic vasospasm or no response to initial medical management, all patients required endovascular treatment. Intra-arterial milrinone resulted in (mean 9 ± 8%) increase in arterial diameter. Four patients (80%) neurologically improved within 1 hour post targeted treatment. No adverse neurological events were associated with intra-arterial milrinone. Duration of continuous intravenous milrinone was 12.4 ± 3.6 days. Two patients remained intubated for >5 days and 1 patient died in ICU. All other patients (80%) had a good neurological outcome (modified Rankin scale ≤ 3) at the 6 -12 months follow-up.

Despite modest angiographic response to targeted intra-arterial milrinone on a background of high dose continuous intravenous infusion, neurological improvement in symptomatic cerebral vasospasm was attained.

Resting energy expenditure (REE) in stroke patients is significantly altered by hypermetabolism and hypercatabolism following an increase in levels of cytokines and counterregulatory hormones. 

Two patients were receiving calories from propofol at the time of indirect calorimetry (1003 and 975 calories in 24 hours). Enteral regimens were decreased to account for propofol calories. The PSU 2003b underestimated energy requirements in 3 patients ; -1501, -443, and -461 calories (50.1%, 22.4%, and 26 .7%) compared to results of the indirect calorimetry.

Indirect calorimetry remains the gold standard to measure energy requirements in hospitalized patients. The PSU 2003b equation underestimates calorie requirements in hypothermic SAH patients compared to measured results of indirect calorimetry. Because indirect calorimetry is not available in all facilities, further research studies should address a corrective cofactor for predictive equations used for patients with hypothermia.

Previous investigators identified an association between intraventricular hemorrhage and neurogenic fever in aneurismal subarachnoid hemorrhage (aSAH). Thus, we hypothesized that patients with neurogenic fever would have higher RBC counts in the CSF than those patients without neurogenic fever.

We conducted a retrospective review of consecutive patients admitted to the Neurological Sciences Intensive Care Unit from January 2011-December 2012 with a diagnosis of aSAH. Neurogenic fever was defined as at least one temperature of 38.4°C in the first 72 hours after hemorrhage with negative cultures. Outcomes measured included length of stay (LOS) and discharge Glasgow Outcome Score (GOS).

Of 71 

Contrary to our original hypothesis there was not an association of CSF RBC count and risk of fever. A higher H&H score was predictive of early fever. Our results suggest that severity of injury may in fact be the driving factor of neurogenic fever rather than the presence of IVH.

Changes in cerebral perfusion pressure (CPP) over time and their association with delayed cerebral ischemia (DCI) remain unclear. We studied the temporal profiles of CPP change over time and tested their association with DCI.

CPP values for 238 subjects were retrospectively obtained. CPP was calculated using intracranial pressure (measured via external ventricular drain) and mean arterial pressure. Values were obtained every 2 hours for 14 days. Induced hypertension was used prophylactically to prevent vasospasm. Growth curve analysis was used to test the linear and quadratic change over time and multivariable logistic regression was used to test the relationship between DCI and proportions of CPP values >110, >100, <70, and <60 mmHg. Each analysis used separate models that controlled for aneurysm treatment method (endovascular coiling vs. surgical clipping) and Hunt and Hess grade. DCI was defined as clinical deterioration attributed to impaired CBF.

Subjects were aged 53±11.4 years, 69% females and 88% Caucasian. DCI was diagnosed in 41.9%. Inter-individual differences contributed to 39% of variation in CPP. A significant linear increase in CPP over time (β =0.06, SE=0.006, p<0.001) was found. The covariance (= -0.52, SE=0.09, p<0.001) between the intercept and the linear change was negative, indicating subjects with low CPP had a faster rate of increase and those with high CPP had a slower rate of increase. The rate of quadratic change (-0.0006) was small compared to the linear change (0.06), showing greater change in the initial interval after admission (until day 5). Each 10% increase in the percentage of CPP>100 and >110 mmHg were associated with 1.21 and 1.43 increase (respectively) in the odds of DCI.

When used prophylactically, induced hypertension contributes to higher CPP values. Based on the CPP trends and correlations in this study, induced hypertension may not confer its expected benefits after aSAH.

Shivering during cooling is a common problem that can hinder its effectiveness. The ability to continuously measure and incorporate a shivering signal into cooling algorithms could facilitate treatment and outcomes for patients. We tested a new method to measure shivering continuously using a series of triaxial surface accelerometers (TSA) placed in locations that correlate with measurement of the bedside shivering assessment scale (BSAS).

This was an IRB approved study of healthy volunteers undergoing cooling for 30 minutes with the Arctic Sun 5000 Temperature Management System. The device was maintained at 4C to maximize the shivering response. BSAS measurements were performed every 5 minutes by two independent raters who remained blinded to each other's assessments and to data from the shiver sensor. TSAs were placed in 3 muscle locations on each subject: the pectoralis, deltoid, and quadriceps. Primary endpoint was to determine whether TSAs could accurately measure shivering as determined by the BSAS. Additional analyses were conducted to determine if the location of TSAs impacted shiver detection correlation and to test an algorithm to discern shivering from nonshivering movements.

The mean age of the 14 volunteers was 33+/-7 years with a mean body surface area of 1.9+/-0.2 kg/m 2 ; six (43%) of the volunteers were women. The median BSAS throughout the cooling period was 2 (range 0-3). TSAs demonstrated an overall good ability to detect shivering with an mean positive prediction agreement of 85+/-15%. Highest kappa agreement was observed with the sensors placed on the deltoid muscle (kappa=0.56); the lowest kappa was observed with the quadriceps muscle (kappa = 0.21). Each instance of nonshivering movement was rejected by the algorithm developed for the TSA.

TSAs placed over specific muscle groups may provide an accurate assessment of shivering frequency and intensity. Additional studies should include further assessment of the signal with measures of energy expenditure in patients in the ICU setting.

Financial Support: N. Badjatia is a scientific advisor for C. R. Bard, Inc., Bard Medical Division.

Therapeutic temperature modulation (TTM) has been proposed as a potential therapy for refractory status epilepticus (RSE). Despite evidence from animal models, few human cases have been reported. Objective: To characterize the effect of TTM in patients with RSE.

We identified all non-cardiac arrest RSE patients admitted to Columbia University between 1/2008-1/2013 who underwent TTM for management of refractory fever, cerebral edema, or seizures. Baseline characteristics, medication and temperature data were collected. EEGs were evaluated blinded to temperature management, and categorized on an ordinal scale of time spent in epileptic activity. 

These preliminary observations support a role for hypothermia in management of RSE, but suggest that normothermia may be ineffective for seizure control. A temperature threshold may exist for control of seizure activity with TTM.

Changes in the levels of inflammatory and anti-inflammatory molecules in pathologies such as acute brain injury, sepsis and post cardiac arrest have been well documented as well as the changes that occur after animals with such conditions have been exposed to controlled hypothermia as a therapeutic maneuver. It is currently unknown how hypothermia by itself may affect the expression of these interleukines. We investigated interleukins gene expression in rats subject to controlled hypothermia in order to better understand the physiological changes associated to isolated hypothermia.

Using real-time PCR, we performed gene expression analysis of IL-10, IL-6, IL-1βand TGFβ-1 in peripheral blood of 11 male rats subject to controlled hypothermia (32 °C, blood drawn at 2 hours after reaching target temperature) and 10 male rats subject to normothermia (37 ° C).

Granting a value of 1 to the results obtained from the control group (normothermia), we were able to compare the results obtained from the experimental group (hypothermia) as follows: IL-10: 1 vs. 0.13, IL-6: 1 vs. 0.82, IL-1β: 1 vs. 2, TGFβ-1: 1 vs. 1.17. These preliminary results indicate a lack of statistically significant differences in the expression of these interleukins between the two groups of rats.

Probably due to the small number of studied animals we couldn't detect a strong difference in the interleukin gene expression in the two groups of animals however we were able to detect a trend towards a decrease of IL-10 gene expression and towards an increase in the IL-1β gene expression in the hypothermia group in this pilot study. Due to the potential clinical applications, further characterization of the physiological responses to isolated hypothermia and rewarming is needed.

Recent clinical studies on therapeutic hypothermia (TH) for traumatic brain injury (TBI) suggest that achieving TH prior to evacuation of intracranial hematomas (ICHs) appears to be beneficial. Early preoperative TH combined with timely surgical intervention is thought to be advantageous in mitigating reperfusion injury following ICH removal. In clinical use of TH, it is reasonable to individualize the TH duration according to the patients' intracranial pressure (ICP).

We retrospectively reviewed the medical records of TBI patients who were treated with TH over 10 days since 2003 in our Level 1 Trauma center. TH was performed by means of surface cooling under general anesthesia. The core temperature was maintained at 33-34 °C at least 2 days and rewarmed 1°C/day or slower when the ICP remained within normal. Enteral feeding and tracheostomy were carried out within 24 hours after arrival. Clinical outcome at 6 months was assessed using the Glasgow Outcome Scale (GOS).

Among 54 TBI patients treated with TH, 9 met the inclusion criteria. Six men and 3 women, mean age of 34.7 years (range 7 to 64) with mean Glasgow Coma Scale on admission of 7 (range 3 to 13, 3 patients deteriorated rapidly), underwent TH between 11 and 19 days of duration. All underwent surgical removal of ICH or contusion, 3 had once and 6 had twice. The mean core temperature reached 34.3°C at the time of initial craniotomy. Their GOS was favorable in 6 (3 good recovery and 3 moderate disability) and unfavorable in 3 (1 severe disability, 1 persistent vegetative state, 1 death). Two patients developed adverse events (1 pneumonia and 1 meningitis) but recovered after early and careful treatment.

Rapid induction and prolonged use of TH together with timely surgical intervention could be feasible and safely achieved with prudent neurocritical care management.

In comatose survivors (CS) of out-of-hospital cardiac arrest (OHCA), it is common to prognosticate neurologic outcomes at 72 hours; however, this practice is based on data preceding the standard use of therapeutic hypothermia (TH). The aim of this study is to compare patients who undergo TH and awaken before vs. after 72 hours. The secondary aim is to describe the incidences of brain death and withdrawal of life sustaining treatment (WLST) in CS-OHCA.

Prospective observational cohort consisting of consecutive adult, witnessed CS-OHCA admitted to an ICU at an academic tertiary-care hospital from July 2007 to December 2011. Data were analyzed using descriptive statistics on neurologic survival of patients who awaken (as defined by following commands).

Of 423 OHCA, 268 (63%) survived to ICU admission; of those, 118 (28%) received TH. Mean age of patients undergoing TH was 57; 78 (66%) were male and 74 (63%) had an initial rhythm of ventricular tachycardia or fibrillation. Sixty-five (55%) patients awoke. Twenty-five (39%) of the 65 patients awoke after 72 hours (vs. 40 before) and 20 of these 25 (31%) (vs. 36 of 40) had good outcomes as defined by Cerebral Performance Category (CPC) 1-2 (p = 0.08). All forty-five of 118 (38%) patients treated with TH who had WLST died after this decision, and 5 (4%) had WLST within 72 hours. Five of 118 (4%) patients progressed to brain death; the remaining 46 deaths were due to hemodynamic, respiratory or infectious causes irrespective of WLST status.

The majority of CS-OHCA treated with TH die following a decision to WLST, and only a minority die from a primary neurologic cause; as many as 40% awaken after 72 hours, many with a good outcome. Prognostication leading to WLST decisions at or before 72 hours appears premature in patients post-TH.

Electroencephalographic (EEG) monitoring is recommended during therapeutic hypothermia (TH), but availability is variable and no specific EEG system or clinical endpoints have been proposed. We report simultaneous monitoring with a simple processed EEG (SPE) including bispectral index (BIS) and suppression ratio (SR) compared to a full montage continuous EEG (cEEG) system.

14 adult cardiac arrest patients were monitored with cEEG and SPE within 6 hours of ROSC during TH in 2012-13. Demographic and clinical data were prospectively entered into the INTCAR registry, and EEG background was determined with cEEG review blinded to BIS/SR data using ACNS 2012 standard Critical Care EEG Terminology. Corresponding BIS and SR were determined blind to cEEG. Data are reported as median (IQR).

Initial rhythm was VT/VF in 36%, asystole in 36%, and PEA in 28%. Time to ROSC was 17 (14-28) minutes, age was 58 (40-72) years, 64% were male. SPE was placed 64 (38-89) minutes earlier than cEEG, which revealed a flat or near flat (F) background in 9, discontinuous (D) in 2, and continuous (C) in 3. The corresponding BIS values were 4 (1-6) in F, 33 in D, and 38 in C. SR values were 95 (92-99%) in F, 12.5% in D, and 25 in C. Of 9 patients with flat background, 1 (11%) made a good recovery. All patients with D or C background in this pilot study made a good recovery.

A simple processed EEG system (including BIS, SR, and frontal raw EEG) appears to reliably identify background patterns determined from continuous full EEG montage monitoring among a pilot sample of cardiac arrest patients treated with TH. Additional research is warranted. 

Hypothermia as a therapeutic modality for refractory intracranial hypertension (RICH) in traumatic brain injury (TBI) patients remains controversial. We present our experience with TH for RICH.

We prospectively studied all severe TBI patients (GCS <9) admitted between June 2010 and January 2013. Patients with RICH ( ICP > 25 mmHg for > 10 minutes despite maximal tier 1 and 2 therapy ) received TH according to our institution's protocol ( goal T = 33°Cfor 48 hours, then rewarming at 0.1°C/hr; if ICP rises during rewarming, protocol is repeated for 48 hr, and so on).

A total of 42 severe TBI patients were enrolled (17 required TH, and 25 did not). Good outcome is defined as GOS of 4-5 at 3-6 months. Characteristics of the TH and no TH (NTH) groups included: average age in years (35 vs 42, p = 0.3), male gender (82% vs 88%, p = 0.9), initial GCS (4.9 vs 5.4, p = 0.6), and initial ICP in mmHg (26 vs 28, p = 0.6). They expectedly differed on the duration of ICP monitoring in days (6.4 for TH vs 3.6 for NTH, p <0.01). There was no difference in good outcome between the 2 groups (41 % in TH vs 56% in NTH, Odds ratio [OR] = 0.9; 95% confidence interval [CI], 0.3 -2.6, p = 0.8). 7 patients in the TH group sustained cardiac arrest (CA) in the field (All died). Good outcome analysis using isolated TBI groups (excluding the TBI/CA group) showed: 70% in TH vs 56% in NTH, OR = 1.5; 95% CI, 0.5 -4.7, p = 0.5).

Pending results from large multicenter studies evaluating the effect of therapeutic hypothermia on ICH and outcome, therapeutic hypothermia should be included in the standard armamentarium of therapies used to control refractory ICP in severe TBI patients.

Fever is common in patients with subarachnoid hemorrhage (SAH) and has been associated with poor outcome. It is treated aggressively in the neuro ICU. In some cases, however, fever is refractory to the mainstay antipyretics acetaminophen and empirical antibiotic treatment. Such cases often require more aggressive normothermia therapy with surface or intravascular cooling. Cyclooxygenase -2 (COX-2) is selectively induced by proinflammatory cytokines at the site of inflammation and may contribute to fever in patients with SAH. COX 2 inhibitor therefore may be effective treatment for refractory fever.

We have conducted a retrospective analysis of all SAH patients who have been admitted to our ICU from October, 2012 to June, 2013. 12 patients were identified to have refractory fever and were treated with COX 2 inhibitor celecoxib for fever control. Temperature data in the 48 hours prior to initiating therapy were reviewed as compared to the period of 72 hours after starting treatment with celecoxib. Primary outcome is the decrease in mean temperatures in the post treatment period as compared to the pre-treatment period.

Of the 12 patients identified, 7 achieved reduction in the mean temperatures in the post treatment period. The average reduction in mean temperatures among this group is 1.38 degree Celsius. 5 out of 12 patients had increase in mean temperature in the post treatment period (average increase 0.48 degree Celsius). Overall, Celebrex use was associated with significant decrease in mean temperatures ( 99.7 versus 100.3) (p = 0.042).

COX-2 inhibitor Celecoxib appears effective for refractory fever in SAH patients. The limitations of this study are its small sample size and retrospective nature. Further prospective controlled studies are warranted to adequately investigate the effectiveness of celecoxib for refractory fever Financial Support: None

Metachromic leukodystrophy (MLD) is an autosomoal recessive lysosomal storage disorder, arising from a deficiency in arysulfatase A, resulting in extensive demyelination of the central and peripheral nervous system. Early onset MLD progresses quickly and often fatal within a few years of onset. Adult onset MLD follows a more benign course and typically presents with progressive dementia and behavioral changes. We describe an unusual case of adult onset MLD in a patient with common variable immunodeficiency syndrome (CVID) presenting rapidly, resulting in death 5 months after his initial presentation.

Case report.

A 27 year old Caucasian man with a history of CVID, on monthly IVIG therapy, presented with rapidly progressing bilateral upper extremity weakness and anartheria. Initial brain MRI showed symmetric minimally enhancing lesions of the bilateral basal ganglia, left thalamus, and a few cortical/subcortical lesions. Infectious, toxic, and paraneoplastic work up was unremarkable. He was diagnosed with ADEM but despite treatment with high dose intravenous steroids and IVIG he progressed to complete spastic quadriparesis and bulbar paralysis over the course of one month and developed respiratory failure two weeks later. Serial brain MRI showed progression of abnormal T2 and flair signal eventually involving all white matter, extending into the deep gray matter structures, pons and medulla. Genetic testing showed two deleterious mutations in the ARSA gene consistent with arylsulfatase A deficiency. Upon diagnosis of MLD his family elected to pursue comfort care nearly five months after symptom onset. He died shortly thereafter.

Although thought of as a more "benign" disease variant, adult onset MLD may have a severe and even fatal presentation in patients with immune dysfunction. It is unclear how immunodeficiency influences the progression of MLD. This case raises questions about the relationship between the immune system and rapid demyelination even outside the boundaries of multiple sclerosis and similar disorders.

Transtentorial herniation (TTH) is a life-threatening syndrome. The pupillary exam is central to its detection. Automated pupillometry provides quantitative measures of the pupillary response including maximal diameter, latency, constriction velocity (CV), minimal diameter, dilation velocity (DV), and a composite Neurologic Pupil Index (NPi TM ). We utilized automated pupillometry to describe the pupillary response in TTH events.

An infrared pupillometer (NPi-100 TM ; NeurOptics, Inc., Irvine, CA) was used to observe the pupillary response in a patient with a large subcortical abscess. Blinded measurements were obtained along with clinical assessments. Data was downloaded and translated via software provided by NeurOptics. Data was processed using Mathematica 8.0 (Wolfram Research, Inc., Champaign, IL).

We monitored a series of 10 TTH events in a single patient who presented with a GCS of 7 and small, round, unequal, sluggish pupils. The first TTH occurred on day 5 after which an ICP monitor was placed. All subsequent events occurred without associated elevation of ICP. During each TTH, maximal and minimal pupillary diameters equalized, consistent with the clinical assessment of an unreactive pupil. A reduction in CV, DV, minimal and maximal diameters were noted to have occurred approximately 24 hours before the patient's first TTH, creating a new baseline, as these measures never recovered after the first TTH. An abnormal NPi was noted to precede 7 of the events, occurring at a median of 5.05 hours prior to TTH. All TTH were reversed following clinical intervention with normalization of the NPi at a median of 34 minutes.

Automated pupillometry provides objective measures of the pupillary response traditionally described by a subjective clinical assessment. Although an abnormal NPi has promise as a precursor to TTH, a formal clinical trial will be needed to further understand these observations.

Financial Support: Pupillometer equipment provided by NeurOptics.

Anti-N-methyl-D-aspartate receptor encephalitis is an autoimmune syndrome that presents with a variety of symptoms, and is commonly associated with mature ovarian teratomas. Most patients with anti-NMDAr encephalitis develop a multistage illness that progresses from psychosis, memory deficits, seizures, and language disintegration into a state of catatonic unresponsiveness with abnormal movements, and autonomic and breathing instability.

A case study will be presented about a patient with this syndrome. An ovariam teratoma was resected. Intravenous corticosteroids, IVIG and PLEX were administered.. Second line therapies included rituximab and cyclophosphamide. Serum and CSF samples were analyzed every six weeks. The patient was frequently monitored on EEG with antiepileptic medication titrated to effect. High dose midazolam and ketamine drips were provided to manage autonomic dyskinesia.

The patient showed no improvement to first line therapy. Initial rounds of rituximab and cyclophosphamide yielded negative side effects. The patient had frequent dyskinetic storms marked by severe tachycardia, tachypnea, pyrexia, and violent dyskinetic movements. EEG findings showed no correlation to the dyskinesia, but routinely demonstrated generalized slowing and multifocal cerebral dysfunction. The patient was weaned off ketamine and midazolam. Subclinical seizure activity continued, and antiepileptic drugs were titrated for effect.

Despite aggressive immunosuppression, the patient remained comatose. The patient did present with a significant reduction in the severity and frequency of dyskinetic storms. The patient remained on a ventilator. Protocols were established for hemodynamic control during dyskinetic storming with increasing amounts of sedation. The patient experienced chronic GI dysfunction. Six months after the final administration of cyclophosphamide, the patient developed hemorrhagic cystitis. The patient went into PEA arrest and required abdominal decompression. Subsequently the patient required multiple abdominal washouts and vasopressor support. Care was withdrawn after a ten month stay in the Neuro ICU. Since these cases are rare, it is important to share the Neurocritical care team's experience.

Reversible cerebral vasoconstriction syndrome (RCVS) is characterized by reversible multifocal narrowing of cerebral vasculature via angiography. We report the first case of ponatinib, a new tyrosine kinase inhibitor (TKI) used to treat chronic myeloid leukemia (CML), associated with RCVS and its possible mechanism.

Case report.

A fifty-two year old female with CML, systemic lupus erythematosus, hypertension and a coiled posterior communicating artery aneurysm one year prior, presented to the hospital complaining of sudden onset of severe headache and diplopia. Ponatinib had been started four weeks prior with a significant increase in her blood pressure and she had required treatment in the emergency room two weeks after initiation for hypertensive emergency. Initial blood pressure was elevated and neurological examination revealed a right internuclear ophthalmoplegia. Magnetic resonance imaging revealed an acute right midbrain infarct with moderate to severe narrowing of the basilar artery and bilateral supraclinoid internal carotid arteries. Computed tomography corroborated these findings and also demonstrated irregularities in the right vertebral and bilateral middle cerebral arteries. The patient was admitted with concern for RCVS and ponatinib held. Cerebrospinal fluid analysis demonstrated no evidence of vasculitis and rheumatological workup showed no abnormalities and no evidence of active lupus. Intracranial velocities were diffusely elevated via transcranial Doppler (TCD). Cerebral angiography showed evidence of fibromuscular dysplasia in addition to severe vasoconstriction of the left anterior cerebral and right internal carotid arteries, which improved with intra-arterial verapamil infusion. Oral verapamil was started before transitioning to oral nimodipine. By discharge she had complete resolution of her headache and neurological deficits with significant improvement of TCD velocities.

This patient's symptoms and temporal relationship with initiation and discontinuation of ponatinib implicate it as the likely causative agent of her RCVS. TKIs increase production of endothelin-1 by vascular endothelium, which has been implicated in the mechanism of cerebral vasoconstriction.

Post-traumatic vasospasm (PTV) is a documented secondary insult to patients with traumatic brain injury (TBI). Its true incidence is unclear, but it has been associated with the development of cerebral ischemia and poor neurological outcome. The current TBI guidelines do not mention screening for or management of PTV to avoid these detrimental effects. Furthermore, as TBI patients are often comatose, the typical signs of cerebral vasospasm may be missed. Thus PTV is frequently discovered after significant injury has already occurred. Identification of PTV with severe refractory hypertension as the presenting sign has not been described in the literature.

Case report and literature review.

We present a case of a 25-year-old male with no past medical history who sustained a TBI in a high-speed motor vehicle crash. Initial GCS was 9 and head CT showed diffuse subarachnoid hemorrhage. On hospital day 3, he had improved neurologically to GCS 15, and on day 4 he developed new hypertension refractory to multiple medications. On day 5, he became progressively less arousable. MRI/MRA and CTA confirmed severe vasospasm of the anterior circulation. The patient was taken to the interventional neuroradiology suite and successfully treated with balloon catheter angioplasty and intra-arterial nicardipine. Hemodynamic augmentation therapy as well as oral nimodipine were initiated and continued over the following several days. Transcranial doppler monitoring was performed daily to follow the Lindegaard ratios and document the degree of cerebral vasospasm throughout his course.

An acute hypertensive response following TBI may be a presenting sign of cerebral vasospasm. As aggressive blood pressure reduction can precipitate cerebral ischemia, PTV should be considered when a TBI patient develops acute refractory hypertension.

The traumatic cerebral contusions is a common clinical entity but its prognosis varies among patients. The frontal location contusions are vulnerable to rapid deterioration and resulting in severe elevated intracranial hypertension. In this clinical situation, the development of non-convulsive seizure may be overlooked by the fully sedated state because of aiming at the control of intracranial presure.

We reported a case of delayed bifrontal contusions resulting in postoperative refractory intracranial hypertension. It is not controlled by therapeutic hypothermia and other medical measures. The continuous electroencephalography showed nonconvulsive status epilepsy exists. The condition was well improved after appropriated anti-epileptic treatment.

The early detection of non-convulsive seizures in patient with refractory intracranial hypertension is prime important. Intriguingly, there are studies demonstrated that therpeutic hypothermia may suppressed the seizure activity. It must be bear in mind that the necessity of continuous electroencephalography monitoring in those paitent if there is unanswered intracranial hypertension exist.

Transcranial Doppler (TCD) and Electroencephalography (EEG) studies were used to follow a patient with cerebral hemorrhage after decompressive craniectomy (DC). Significant increases in Flow velocity (FV) ipsilaterally to the site of Decompressive craniectomy and a bilateral decrease of the Pulsatility Index (PI) has been related with changes in cerebral blood flow and outcome. Also it is well described that a progressive disappearance of fast waves occurred when the cerebral blood flow falls.

A 65-year-old male with a right temporo parietal DC with evacuation of hematoma was admitted to the Critical care Unit. Thirty days after the patient was awake and started following simple commands with aleft hemiparesis sequelar. TCD findings Immediate postoperative TCD showed bilateral MCA with increased PI (1.2 and 1,3) and significant asymmetric mean FV (53 and 70 cm/s) was seen in the middle cerebral artery right and left respectively. After the first week we observed a return to normal PI and higher FV. EEG Findings The first non-continuous EEG performed on the third day of evolution showed a core activity with predominant beta rhythm an isolated delta activity without paroxysmal discharges. Seven EEG studies were arranged according to the International 10-20 system using 24 recording electrodes. The brain electrical activity on the side of the decompressive craniectomy followed the electrical recovery evidenced on the left hemisphere. The temporal evolution showed a predominantly beta activity followed by a diffuse theta delta activity in the left hemisphere and a slower recovery in the right hemisphere with the progression from isolated delta to delta predominant activity.

Incipient normalization electrical activity preceded normalizing of FV and PI but these also preceded clinical recovery.

Heparin-induced thrombocytopenia (HIT) is a life threatening thrombotic complication of the use of unfractionated or low molecular weight heparin. Recommended treatment involves cessation of heparin and initiation of anticoagulation with a non-heparin agent; however, when anticoagulation is contraindicated, plasma exchange may be effective in clearing antibodies directed against heparin-platelet factor-4 (heparin-PF4) complex and reducing the risk of thrombosis.

A 64 year old woman who sustained a T12 burst fracture after a 10 foot fall was admitted to our ICU. During her hospital course, she developed paroxysmal atrial fibrillation. A few days after her T10-L2 posterior spinal fusion, she became unresponsive and developed thrombocytopenia due to HIT. CT of the head revealed bilateral occipital hypodensities with right occipital ICH. She was also acutely hyponatremic. She was diagnosed with adrenal insufficiency due to bilateral adrenal hemorrhage. Given the severity of patient's condition and initial high optical density (OD) value of her heparin antibody using ELISA measurements, there was concerned for expansion of ICH and additional thrombotic events so plasma exchange was employed. We obtained daily heparin antibody ELISA measurements as well as heparin-induced platelet aggregation (HIPA) assays on both pre-and post-exchange samples with the goal of terminating plasma exchanges when HIPA testing became negative and/or the patient was unable to tolerate the procedure. To minimize risk of additional hemorrhages, we obtained coagulation and fibrinogen levels daily. We used FFP as the replacement fluid for the procedure.

The patient underwent five plasma exchange sessions. HIPA assay and anti-heparin antibody levels became negative after the fourth exchange session. Bivalirudin was initiated for anticoagulation nine days after ICH. Platelet levels trended up after the initial platelet transfusion.

This case illustrates the utility of plasma exchange in rapidly removing anti-heparin antibodies in acute HIT, particularly when anticoagulation is contraindicated.

Spontaneous intracerebral hemorrhage (ICH) in end stage liver disease (ESLD) patients with coagulopathy of liver disease (CLD) is presumed to be due to lack of adequate hemostasis. We hypothesize however, that in some cases venous hypertension can occur because of nasal packing. This may increase the venous pressure within the "danger triangle" of the face leading to intracranial venous hypertension and possibly make them more susceptible to small vessel deep ICH, similar to patients with sagittal sinus thrombosis.

We report a small case series of two ESLD patients with CLD, epistaxis treated with nasal packing and spontaneous ICH.

Two female, ESLD patients, aged 52 and 58 with CLD and epistaxis requiring nasal packing were found to have ICH. Both patients had elevated venous pressure with CVP's ranging from 9-23mmHg. The patients had GCS scores of 3 and6 (E1M5V1T) respectively at the time of ICH discovery. In both patients CT of the head showed ICH in the subcortical frontal white matter. CLD was reversed in both patients with blood products and desmopressin prior to and following discovery of ICH. Attempts were made to decrease venous pressures in both patients using negative fluid balance via continuous renal replacement therapy. The first patient died hospital day 16 due to progressive ICH and progression to brain death. The second patient also progressed to brain death and died on hospital day 21.

These two cases illustrate patients with subcortical frontal ICH in the setting of CLD and nasal packing for epistaxis. Decreasing the venous pressure may help reduce epistaxis and bleeding from nasal veins in this region and thus prevent severe venous hypertension exacerbated by nasal packing. Further studies are needed to explore this hypothesis and understand how to appropriately correct CLD in critically ill liver patients.

Fat embolism is a serious and potentially devastating complication usually associated to hip or long bones fractures. Neurological complications of this clinical picture are generally due to capillary damage (secondary to endothelial injury and DIC) causing small infarctions and microhemorrhages. We present a case of a large intracranial vessel occlusion due to paradoxical fat embolism that was successfully recanalized.

Single case report.

80 years old male that developed acute onset of severe left-sided weakness upon emergence from general anesthesia of an uncomplicated hip replacement surgery. Non-infused head CT showed normal brain parenchyma and a "Hypodense right MCA" that involved the M1 and M2 segments suggesting the presence of either air or fat tissue located inside of the artery. Digital substraction angiography confirmed the occlusion of the proximal right MCA; a Solitaire ® stent retriever device was positioned and a small white "mass" was retrieved achieving recanalization of the previously occluded vessel and normal distal flow in the right MCA territory (TICI 3). Despite reperfusion, the patient developed a moderate size right MCA infarction with some degree of asymptomatic hemorrhagic transformation. The pathological examination of the specimen obtained from the MCA showed bone marrow fat tissue. A transesophageal echocardiogram showed an atrial septum defect with substantial right to left shunt at rest.

Bone marrow fat tissue can cause paradoxical large vessel embolic stroke in the context of bone trauma, especially in the elderly when bone marrow has been replaced by fatty tissue. We present a case of a tomographic "Hypodense MCA sign" secondary to fatty occlusion of the intracranial vessel. Endovascular fat retrieval or liposuction can be successfully achieved in these circumstances with the new generation of intravascular devices.

In general, pediatric stroke leads to significant morbidity (66%) and mortality (25%). The outcome in the event of the large artery occlusion, especially, internal carotid artery terminus (ICA-T) is not known. There is no approved intravenous thrombolytic therapy and only few case reports of intra-arterial thrombolysis (IAT). Mechanical thrombectomy is not primarily offered the lack of data and fear of possible cerebral vasospasm associated with device. .

The patient was an 8-year-old boy with transposition of the great arteries, spstein anomaly of the tricuspid valve, and status post correction of co-arctation of the aorta along with VSD closure who presented to the hospital with acute stroke with NIHSS 11 and CT revealed a dense right MCA sign. Patient underwent endovascular therapy within 3.5 hours after an informed consent from parents Cerebral angiography demonstrated internal carotid artery (ICA)terminus occlusion in which thrombus extending from the right ICA into the right middle cerebral artery(MCA) and anterior cerebral artery (ACA).

To prevent vasospasm associated microcatheter manipulation, intra-arterial nitroglyserine drip was administered through the guiding catheter. The thrombus was resistant to IAT. However, no cerebral vasospasm was noted during catheter manipulation. Therefore, a thrombectomy device (Solitaire 4x20 mm) was used to remove clots resulted in complete recanalization and the NIHSS reduced to 6. MRI day 2 and 7 demonstrated restricted diffusion in the ACA and MCA territory without hemorrhagic transformation. 90 days outcomes; NIHSS 2 and mRS 1.

Emergent endovascular thrombectomy in acute pediatric ischemic stroke patient using Solitaire device may results in recanalization of the large cerebral artery that is resistant to intra-arterial thrombolysis with good clinical outcome.. The thrombectomy device may not induce cerebral vasospasm if intra-arterial vasodilator is administered through guiding catheter during thrombectomy procedure. This is the first thrombectomy procedure in pediatric patient in our knowledge and further studies are required Financial Support: None

Aberrant communication between a blood vessel and air, typically introduced through surgical, traumatic or diagnostic means, may produce an air embolism, which can lead to dire neurological consequences.

A 58 year old female with a history of atrial fibrillation status post cardiac ablations one and four weeks prior to admission presented with acute onset weakness and numbness of the left arm and leg. Symptoms resolved within 30 minutes prior to reaching the ER. CT brain and CT angiogram of head and neck did not show any significant abnormalities. Later on the day of admission, the patient developed left sided hemiplegia. A second CT head was suspicious for air in multiple arteries of the right frontal, parietal and occipital lobes. A follow-up MRI brain revealed acute infarction of both hemispheres and the left cerebellum, with the majority of infarction in the right fronto-parietal lobes. Due to multi-territorial findings and worsening neurological exam, we suspected a persistent source of emboli to the brain. An emergent echocardiogram showed no shunts. This was followed by a CT chest, and no evidence of fistulae was found. The patient's mental status continued to deteriorate, and her exam became consistent with bi-hemispheric lesions. A subsequent Cardiac CT Angiogram suggested a fistula between the distal esophagus and the left atrium with air bubbles in the left ventricle.

This is a case of multi-territorial stroke due to an air embolism introduced through an atrio-esophageal fistula after cardiac ablations. Medical literature has noted multiple avenues for introduction of air emboli into the brain such as cardiac ablation, lung biopsy, venous catheter placement, upper gastrointestinal endoscopy with high air pressure, and positivepressure maneuvers employed in cardiac resuscitation. Our case is unique compared to others reported in literature as it represents a late complication of cardiac ablation therapy. 

This is a case report. A 68-year-old man with recent pneumonia presented with decreased level of consciousness, left hemiplegia and neglect.

Initial head computer topography (CT) demonstrated dilated and contrast enhancing right lateral ventricle and 2 cortical ring-enhancing lesions. Intravenous antibiotics were administered. External ventricular drain (EVD) was placed in the right frontal horn, and cortical lesions were surgically removed. Streptococcus viridian was cultured from cerebral spinal fluid (CSF). CSF flow dynamics remained impaired requiring continuous volume drainage via EVD. Multiple head CTs showed isolated dilation of the right temporal horn; however, communication was demonstrated with CT ventriculogram using intrathecal contrast dye. During an EVD clamping trial, severe hypotension and bradycardia developed. Hemodynamic status was restored with vasopressors. Hypotension was extensively investigated, and no alternative diagnosis was identified. Head CT showed worsening dilation of the right temporal horn with compression of the right insular cortex. CSF was actively drained via right frontal EVD, and hemodynamic stability resulted. Shock returned when the EVD was later removed. EVD was replaced via the temporal approach into the right temporal horn. Vasopressors and EVD were weaned, and the patient was discharged.

Bilateral insular cortices are responsible for the regulation of the autonomic nervous system. The right insula cortex controls sympathetic tone, and the left insular cortex mediates parasympathetic tone. Neurogenic shock should be considered when the insular cortex is involved. There are no previously reported cases of neurogenic shock neither from any infectious etiologies nor from isolated ventricular enlargement. This case clearly demonstrates both. Additionally, the treatment for trapped ventricles is limited. This case demonstrates a unique and successful approach with direct placement of EVD into the temporal horn.

Propylene glycol (PG) is a solvent used in various intravenous (IV) medications: lorazepam, phenytoin, etomidate, phenobarbital, and pentobarbital. PG accumulation can induce toxicity with initial manifestations of hyperosmolarity and lactic acidosis, rapidly followed by multi-organ failure. Because of its clinical similarity to sepsis, the diagnosis and treatment of PG toxicity may be delayed. We report a case of PG toxicity in a patient with refractory status epilepticus (RSE) to highlight how PG toxicity can mimic sepsis, and to suggest a solution for early recognition.

Case Report.

A 46-year-old man with intractable epilepsy was hospitalized with a traumatic subdural hematoma, complicated by RSE. Electroencephalographic burst suppression was not achieved despite continuous propofol and frequent IV lorazepam, phenobarbital and phenytoin boluses. Burst suppression was achieved by a pentobarbital load followed by continuous infusion. To facilitate weaning off pentobarbital, IV lorazepam and IV phenobarbital with boluses were continued to achieve goal levels. Within 48 hours after starting pentobarbital, the patient developed lactic acidosis (LA level 14mmol/L) with anuric acute renal failure, liver failure and shock. Broad-spectrum antibiotics were administered after cultures were obtained. Blood, sputum and cerebrospinal fluid cultures remained negative. PG toxicity was suspected. The osmolar gap was 26mOSM/kg. PG level was elevated at 210mg/dL(normal <5mg/dL). IV phenobarbital and pentobarbital were stopped and continuous veno-venous hemofiltration was initiated. The family eventually withdrew care.

High doses of pentobarbital, and multiple repeated boluses of IV lorazepam and phenobarbital led to the accumulation of PG. PG toxicity was confirmed with elevated osmolar gap and PG level. Treatment consists of stopping the offending agents, and renal replacement therapy. Commonalities of sepsis and PG toxicity are lactic acidosis and multi-organ failure, but not hyperosmolarity with elevated osmolar gap. Measuring the osmolar gap is a readily available, inexpensive test, which can lead to timely and appropriate treatment.

Financial Support: None

The incidence of neurogenic stunned myocardium (NSM) in status epilepticus (SE) patients in unknown. We sought to determine the prevalence of NSM in our SE patient population.

Two index patient cases were identified and described. A retrospective review was performed of 100 consecutively admitted patients with the ICD-9 code assigned for generalized SE at our institution from October 2010-November 2011. Patients with anoxic or cardiac arrest events were excluded.

32yo female admitted for a generalized seizure. EEG showed diffuse slow waves with alternating periods of delta slowing and theta waves. She was tachycardic to 170's, BP 60's/40's, unresponsive to intravenous fluid boluses. Nor-epinephrine was initiated to support MAP>65mmHg. Troponin peak 1.45, CK 15,353. Transthoracic echo (TTE) revealed EF 15% with severe global left ventricular(LV) dysfunction with sparing of the basal segments. A pulmonary artery catheter was placed and initial cardiac index (CI) was 2.1L/kg/min. Inotrope (Dobutamine) was required to support CI>2.4L/kg/min. Her CI returned to normal and she was successfully weaned off the inotrope after 4 days. 64yo female was found unresponsive. EEG showed non-convulsive SE. Catscan thorax obtained secondary to hypoxia and tachycardia (HR 148) which demonstrated diffuse bilateral pulmonary emboli and aspiration pneumonia. She developed hypotension requiring Norepinephrine to support MAP>65. TTE revealed EF 40% with basal to mid-septal akinesis, normal anterlateral/inferolateral/periapical contraction, and hypokinesis of remaining walls. TTE repeated nine days later had resolution of her wall motion abnormalities. 100 patients: mean age was 56years old, 40/100 (40%) were female, 56/100 were white. TTE was performed in 43/100 patients. Typical NSM pattern was found in 1/43. Other patterns of wall motion abnormality were found in 6/43. LV function was hyperdynamic in 6/43.

NSM is present in SE patients at an unknown frequency. The clinical effects can be dramatic when these two co-exist. Further epidemiologic study is needed..

Various kinds of neurological illnesses and emotional stress may cause left ventricular dysfunction known as neurogenic stunned myocardium. Subarachnoid hemorrhage, ischemic stroke, major head trauma, and seizure are well-recognized causative illnesses, however, spinal cord lesion has been seldom reported. Takotsubo-like wall motion abnormality is wellknown but focal hypokinesis mimicking acute myocardial infarction has been also reported. We present a case of intramedullary spinal cord hemorrhage presenting with focal cardiac hypokinesis.

A-71-year old woman collapsed immediately after she felt left arm numbness and pain. At the initial hospital where she was transferred, she complained of an anterior chest pain and left arm dullness. An electrocardiogram showed ST segment depression in leads II, III, aVF, and V4-6. A diagnosis of unstable angina was made and nitroglycerin administration relieved the pain. But shortly after she complained of a headache and fell into coma. In an ambulance to our hospital, the SpO 2 was 93% despite forced ventilation and the blood pressure was unable to be measured. Her Glasgow Coma Scale was scored 10 (E4, V1, M4). An emergency coronary angiography was undertaken with vasopressor administration following endotracheal intubation and brain CT scan which detected no lesion. No significant stenosis was detected in coronary arteries despite left ventriculography revealed anteroseptal hypokinesis. When she regained her consciousness next day, tetraparesis became evident. Spinal CT and MRI disclosed intramedullary spinal cord hemorrhage from the medulla to the first lumbar level. Spinal angiography detected no vascular anomaly. Her cardiac function recovered but she remained in tetraparesis and required prolonged ventilatory support for poor spontaneous ventilation.

A case of neurogenic stunned myocardium associated with idiopathic intramedullary spinal cord hematoma was presented. Chest pain radiating to the left arm and anteroseptal hypokinesis made the diagnosis confused.

After primary infection with varicella zoster virus, it remains dormant in sensory nerve roots for life and can have an eclectic presentation with meningoencephalitis, cranial nerve palsies, vasculopathy etc. Here we review a case of fulminate necrotizing rhomboencephalitis as a rare complication of varicella zoster infection.

A 52 year old East African woman with a past medical history of SLE, stage IV lupus nephritis on immunosuppressant therapy was admitted for altered mental status. During the hospitalization, she developed tender vesicles in the lower extremities suggestive of disseminated VZV in conjuction with acute painless right eye vision loss highly suspicious for VZV retinal necrosis. IV acyclovir, andintravitreal injection of ganciclovir were initiated. Patient was intubated after becoming lethargic with complex partial seizures. In the setting of her geographic background and immunocompromised state both antifungals and antituberculosis medications were initiated. Despite being aggressive, patient continued to decompensate from initially have some intact brainstem reflexes to eventually losing all brainstem function. Brain biopsy was deferred due to coagulopathy and thrombocytopenia.

Initial Cerebrospinal fluid initially revealed WBC 68 cells/mm3, segmented neutrophils 85%, lymphocytes 9%, macrophages 1%, eosinophils 0, monocytes 5%. RBC 595 cells/mm3, glucose 8 mg/dl, protein 129 mg/dl. Grams stain and culture were negative while the CSF VZV PCR was positive. Repeat cerebrospinal fluid studies showed WBC 51, segmented neutrophils 90%, lymphocytes 7%, macrophages 0, eosinophils 0, monocytes 3% RBC 15075, protein 244, glucose 25. Cultures of the skin lesions were VZV DFA positive while the viral PCR for VZV was negative.Blood, urine and sputum cultures remained negative. MRI revealed multiple embolic appearing infarcts in multiple vascular territories as well as basilar leptomeningeal enhancement consistent with meningitis.

In patients who are immune compromised, herpes zoster ophthalmicus can rapidly disseminate to fulminant and often fatal encephalitis.

The cerebral artery dissection was one of serious disease leading to stroke. The patients with dissection usually have the severe headache. The medication for relief of headache in these patients remained unclear. Here, we reported one case, with severe headache, that was controlled after beta-blocker.

Case Description 39-aged man had severe headache of sudden onset. Brain MRI and digital subtraction angiography showed the multiple dissection on both ICA and left vertebral artery without subarachnoid hemorrhage or cerebral infarction. Intravenous infusion of nicardipine and hepain was tried and the blood pressure was controlled below 140/90 mmHg.

The severity of headache did not improved over three days, in spite of NSAIDS. Therefore, the carvediol was tried orally by 25mg daily and then headache disappeared. The pulse pressure was reduced from 60mmHg to 35mmHg after taking carvediol. There were no abnormal findings of ophthalmology, dermatology, cardiology and orthopedics. The patients refused the genetic examination for the dissection.

The beta-blocker seems to be the effective medication for the relief of headache in the patients with cerebral artery dissection. The reduction of pulse pressure by beta blocker was possible mechanism for headache relief. We thought that headache was related to the progression of dissection. Therefore, beta blocker could be the candidate for prevention of the dissection progression.

Complex clinical trials in the intensive care units (ICUs) at our urban, academic hospital often rely on bedside nurses, rather than designated research nurses, to follow complex protocols and implement high-risk interventions. Competent delivery of high-risk study interventions requires that ICU nurses understand study purpose, informed consent, risks and benefits, and rigor in intervention implementation and data collection. The need for high functioning, beside research nurses presented us with an opportunity to partner with medical researchers to formally develop the role of bedside research nurse champion. We wanted to enhance participant safety and data quality and provide a mechanism for disseminating research concepts to nurses through training, hands-on experience, and mentoring opportunities. The goal is to foster a multidisciplinary collegial research culture where nurses contribute to study design prior to IRB submission.

Staff nurse champions were identified for two trials in stroke patients that involved complicated interventions. Each champion reviewed the study protocol and participated in specialized training provided by the primary investigator.

Champions met with fellow nurses, leadership, and patients' multidisciplinary teams to develop care plans that incorporated study requirements. They assisted in developing unit-based documentation tools, authored resource materials, conducted staff inservices, and mentored nurses caring for the patients.

Both studies with a champion had improved outcomes. One study had 100% nurse compliance in glucose and daily MRI testing and documentation. The first patient in the second study had no nurse-dependent protocol deviations.

Formalizing the role of the unit-based nurse research champion may foster collaborative research.

Hashimoto's encephalitis (HE) is an autoimmune, neurological disorder with multifaceted clinical presentations associated with high titers of anti-tpo antibodies (TPOa). While the diagnostic aspects of HE remain controversial, several variants of HE have been described. Clinical manifestations of HE include encephalopathy, seizures, behavioral symptoms, involuntary movements, and coma. Although immune-modulating therapies such as steroids, plasma exchange (PE), IVIG and immunosuppressive drugs are routinely used, many patients are refractory to treatment.

We report a case of HE resulting in recurrent seizures refractory to antiepileptics and immune-modulating therapies , which responded to total thyroidectomy.

A 60-year old woman was admitted with status epilepticus (SE). Laboratory evaluations revealed a normal TSH, CSF studies, high TPOa (51 IU/ml) and anti-thyroglobulin antibody (111 IU/ml). Brain MRI showed bilateral hyperintensities of the limbic system. SE was aborted initially with Dilantin, Keppra and Steroids. Subsequently, she had multiple ICU admissions with status epilepticus, epilepsia partialis continua and non-convulsive SE. The seizures remained uncontrolled on 5 anti-epileptic medications. Repeated treatments with steroids, plasmapheresis, IVIG and Rituximab failed to reduce the seizure frequency or the TPOa level. Following total thyroidectomy our patient had remarkable sustained improvement in her symptoms and TPOa levels. On 6 months follow up, she remained seizure free on 2 antiepileptics.

We describe the clinical profile and successful treatment of recurrent SE secondary to HE by thyroidectomy in a patient resistant to immune-modulating treatments. We propose that total thyroidectomy should be considered as a treatment option in HE, when refractory to medical treatment. The long-term benefits of this treatment need to be evaluated further.

Magnetic resonance imaging (MRI) abnormalities isolated to the splenium of the corpus callosum represent a unique radiologic phenomenon, for which various etiologies have been posited. However, the association of vasospasm with splenium lesions has rarely been described. Herein, we present a patient with pituitary apoplexy resulting in a transient lesion of the splenium secondary to vasospasm from pituitary adenoma hemorrhage.

Case report, with review of clinical, laboratory and radiologic data.

A 50-year-old-man presented with complete vision loss of the right eye, associated with worsening right-sided ptosis and a severe headache for two weeks. Neurologic examination demonstrated cranial nerve II and III palsies and an afferent pupillary defect of the right eye. Visual acuity in the left eye was 20/70, with a temporal hemianopia; the right eye could not perceive light. Laboratory evaluation was unremarkable. MRI of the brain revealed an enhancing sellar mass, and signal loss on gradient echo sequencing, suggestive of pituitary apoplexy. A nonenhancing lesion of the splenium was also noted, which demonstrated diffusion restriction with ADC correlate. MR spectroscopy was unremarkable. MRA and CTA of the head were suggestive of vasospasm. This was confirmed by digital subtraction arteriography, which confirmed the presence of multifocal vasospasm. Nimodipine and dexamethasone were started. Visual acuity improved, with resolution of temporal hemianopia of the left eye. The patient underwent tumor resection. Pathology was supportive for pituitary adenoma. Post-operative MRI demonstrated resolution of the splenium diffusion abnormality, and a decrease in size of the sellar mass. The patient was discharged home in stable condition.

Vasospasm secondary to pituitary apoplexy is a rare occurrence which should be treated quickly, as prognosis can be favorable with early management. Sequelae such as isolated splenial lesions should be recognized as a possible consequence, prompting further evaluation for vasospasm.

Influenza has been associated with many neurologic complications including encephalitis, aseptic meningitis, Reye's syndrome, transverse myelitis, GBS, and encephalitis lethargica. These complications are seen more frequently in children and in infections with influenza A, with the exception of Reye's syndrome which is more common with influenza B. Influenza A has also been rarely associated with retinitis. We are presenting a case of rare neurologic complications of influenza B in a female with basal ganglia encephalitis and autoimmune retinitis.

Single observational case presentation with literature review.

We describe a 61 year old female presenting with profound encephalopathy found to have Influenza B. MRI revealed symmetrically increased T2 signal with mild diffuse enhancement in bilateral basal ganglia. As encephalopathy resolved patient developed first diplopia then loss of vision and was found to have autoimmune retinitis. MRI showed decreased intensity following plasmapharesis, and her vision has slowly improved following steroid taper.

Influenza can present with an array of neurologic complications. This case highlights that although encephalitis is more common with influenza A, it can present in influenza B. Additionally, autoimmune retinitis has not been reported in association with influenza B, but is a treatable complication that should not be overlooked.

West Nile virus encephalitis (WNVE) is becoming increasingly common in the United States. Radiologic findings suggestive of WNVE have not been clearly established. False negative WNVE serologic results are common and make distinctive imaging features important clinically. We sought to identify a radiologic pattern typical of WNVE.

Clinical and magnetic resonance imaging (MRI) findings of two patients with WNVE are reviewed. Both patients were immunosuppressed for previous organ transplant and were prospectively identified during admission to a neurospine intensive care unit.

Imaging studies in both patients revealed symmetric T2 and fluid-attenuated inversion recovery (FLAIR) signal hyperintensity in bilateral thalamic nuclei. In patient 1 these abnormalities extended into the hippocampi, cerebellar vermis, medial cerebellar hemispheres, and inferiorly through the brainstem to the distal cervical spinal cord. In patient 2 these abnormalities extended into the corona radiata, and inferiorly into the midbrain. Initial cerebrospinal fluid analysis (CSF) in both patients demonstrated a lymphocytic pleocytosis, consistent with viral infection. Both patients initially had negative enzyme-linked immunosorbent assays (ELISA) for West Nile virus IgG and IgM, with seroconversion during their hospital admission.

Inflammation predominantly affecting bilateral thalamic nuclei may be an indicator of WNVE. When initial serologic studies are negative and this imaging pattern is present, further diagnostic testing should be considered.

Early diagnosis and treatment of hypertensive encephalopathy leading to posterior fossa edema and resulting obstructive hydrocephalus is infrequent. We present a case of obstructive hydrocephalus from hypertensive emergency that was diagnosed early and treated with aggressive blood pressure (BP) management resulting in excellent outcome.

A 46 year old man with known chronic renal insufficiency and poorly controlled hypertension presented to an outside hospital with shortness of breath and BP 265/148 (mean arterial pressure, MAP 186) mmHg. His mental status rapidly deteriorated shortly after requiring intubation. Non-contrasted head CT demonstrated hydrocephalus, cerebellar/brainstem edema and effaced fourth ventricle when he was transferred to our intensive care unit. Neurological examination upon arrival was positive for brainstem function and right upper extremity localization only. Aggressive BP management using nicardipine infusion for MAP 120-130 mm Hg was initiated. Emergent external ventricular drain (EVD) placement by neurosurgery revealed an opening pressure of 26 mmHg. Same day non-contrast brain MRI showed confirmed the CT imaging. A diagnosis of reversible obstructive hydrocephalus due to hypertensive encephalopathy (ROHH) was made and targeted BP control was continued. Following drainage of 174 cc cerebrospinal fluid, EVD stopped working within 24 hours. Another EVD placement was not required as patient's neurological exam continued to improve. Patient was extubated next day and he was back to his neurological baseline on day 4 when his repeat brain MRI showed reversal of posterior fossa edema, hydrocephalus, and 4 th ventricular obstruction.

Hypertensive encephalopathy can be associated with severe vasogenic edema within the posterior fossa precipitating obstructive hydrocephalus. Although ventriculostomy might be required for acute management, early diagnosis of ROHH and aggressive BP management in such patients is crucial to relieve the vasogenic edema and reverse the obstructive hydrocephalus.

Isotretinoin is the mainstay therapy for severe and nodulocystic acne. Myalgia and muscle stiffness have been reported in 16-51% of patients treated with isotretinoin , while elevated serum creatine kinase (CK) levels have been found in up to 41%. In this paper, we report, for the first time, the occurrence of pelvic girdle myopathy in a teen who received isotretinoin therapy.

This was a case study conducted at King Khalid Hospital.

Histopathology and other laboratory data were consistent with an inflammatory myopathy based on a vasculitic process.

Our data indicate thatthe myopathy caused by isotretinoin is an inflammatory process based on vasculitis, resulting in ischemic muscle necrosis.

Development of fulminant hepatic failure is associated with significant risk of mortality, especially in patients that experience cerebral edema and intracranial hypertension. Many patients with intracranial hypertension secondary to hepatic failure succumb to cerebral herniation prior to receiving definitive therapy: liver transplantation. Intravenous indomethacin has been identified as a potential option for patients who have intracranial hypertension refractory to standard therapy. In this case report, we document the use of IV indomethacin for urgent treatment of intracranial hypertension in the United States.

A 42-year-old female transferred to our institution from an outside hospital after presenting with acute liver failure secondary to acetaminophen toxicity. Her mental status deteriorated quickly and, in setting of suspected hepatic encephalopathy, her head CT showed development of cerebral edema. Despite aggressive management with hyperosmolar therapy, lactulose, rifaximin, propofol, hypothermia, and paralytics, she developed refractory intracranial hypertension. In an attempt to decrease her intracranial pressure (ICP), indomethacin 10mg IV push (IVP) was administered.

As indomethacin was administered via slow IVP, the patient's ICP decreased from 30mmHg to 10mmHg almost immediately. This allowed her to remain supine for head CT as part of her transplant work up. ICP remained at goal for approximately 4 hours and 43 min after administration. To finalize her transplant work-up, she required an additional CT on the following day. One additional dose of indomethacin 10mg IVP was repeated, facilitating completion of necessary imaging. She underwent successful liver transplantation on hospital day 6. Postoperatively, despite transplantation she continued to experience intracranial hypertension, as well as multiple additional complications. On post-op day 12, the patient expired secondary to cardiovascular collapse.

Despite the poor overall outcome of our patient, indomethacin 10mg IVP successfully reduced refractory intracranial hypertension, facilitating hepatic transplantation work-up, without documented adverse effects. Financial support: None Financial Support: None

Cerebral edema occurs in 80% of severe hepatic encephalopathy (HE) cases. Head CT for HE has limited sensitivity for cerebral edema and elevated intracranial pressure. We demonstrate these limitations and the potential exacerbation of cerebral edema by continuous veno-venous hemofiltration (CVVH).

We reviewed medical records, including Glasgow Coma Scale (GCS) and serum laboratory values, for a patient presenting with acute liver and renal failure and severe HE. Cerebral edema was quantified by measuring lateral ventricular volumes on serial head CT studies using semi-automated software (Analyze Direct, Mayo Clinic, Rochester MN).

Ventricular volume three weeks prior to admission, with GCS 15, was 62.8mL (ammonia 50 μmol/L). After admission for encephalopathy, when GCS declined to 7T, ventricular volume had decreased to 51.3mL (ammonia 197μmol/L). CVVH was initiated for hyperammonemia and uremia. After 31 hours, ammonia decreased to 78μmol/L while measured serum osmolality decreased from 342mOsm/L to 295mOsm/L. This was accompanied by further decline in ventricular volume (43.1mL) and GCS (4T). Following two weeks of supportive care, ammonia (25 μmol/L) and GCS (11T) improved and ventricular volume (69.7mL) increased. Attending radiologists interpreted all CTs as unremarkable.

Even in suggestive clinical scenarios with prior imaging, head CTs are insensitive for detecting cerebral edema in patients with hepatic failure. Decreases in lateral ventricular volume may reveal worsening edema when imaging is otherwise unremarkable. Patients with hepatic and renal failure may develop a substantial osmolal gap, and rapid reduction of osmolality with CVVH may produce blood-brain osmolar gradients and precipitate cerebral edema. Such patients may benefit from gradual reduction in osmolality. Quantitative lateral ventricular volumes to detect cerebral edema and gradual reduction in serum osmolality to prevent cerebral edema in at-risk patients should be formally investigated.

Tension pneumocephalus is a neurologic emergency requiring prompt diagnosis and treatment. We present a case with unique images of fatal tension pneumocephalus secondary to a ventriculo-bronchial fistula resulting from remote trauma along with a review of the current literature.

year old male was transferred from outside hospital for management of new onset refractory convulsive status epilepticus. Previous medical history was significant for a stab wound to his chest in 1999 that penetrated his left ventricle. Due to the development of recurrent hemoptysis despite surgical repair, a ventricular-bronchial fistula was suspected. Approximately ten years following initial injury, the hemoptysis had failed to resolve leading to a bronchial artery embolization after a ventriculogram failed to show a fistulous connection. The hemoptysis had still failed to resolve and the patient refused further testing or intervention. On admission, he had a prolonged convulsion prompting intubation and transfer to a tertiary care center. Initial CT brain showed scattered pneumocephalus and follow up scan six hours later showed worsening. With tension pneumocephalus thought to be causing the refractory status epilepticus, the patient underwent emergent hyperbaric oxygen therapy (HBOT). Near the end of the HBOT, the patient developed hemodynamic instability and repeat CT brain showed worsening pneumocephalus with diffuse cerebral edema and transtentorial herniation. Emergent echocardiogram showed multiple gas bubbles traversing the left ventricle to the aortic valve. Despite aggressive measures, the patient continued to deteriorate and was pronounced brain dead shortly.

The current case emphasized the importance of documenting the source of air in pneumocephalus. If the suspected source is from the lung, then HBOT should not be performed until the fistula is repaired. In cases where additional respiratory support is required (i.e. status epilepticus), a careful balance should be maintained by limiting positive pressure ventilation and monitoring pneumocephaleus.

Reversible brain stem hypertensive encephalopathy (RBHE) is characterized by transient edema involving the brainstem structures, in the setting of hypertensive crisis. We report two cases of RBHE presenting with obstructive hydrocephalus, requiring external ventricular drainage (EVD).

Retrospective review of patient's charts admitted to the Neuro-intensive care unit at our hospital was performed. We identified two cases of RBHE with obstructive hydrocephalus and their management.

First patient was a 43 year old African American woman with acute onset of confusion, lethargy, slurred speech following a seizure. BP on admission was 258/157 mm Hg. Head CT W/O contrast showed a large hypodensity in the mid brain, pons and the bilateral periventricular regions with effacement of fourth ventricle and basal cisterns. An emergent EVD was placed and the opening pressure was 33 cm h20. MRI showed FLAIR hyperintensities involving the pons, adjacent cerebellar white matter, bilateral temporo-occipital and periventricular areas. Symptoms gradually improved with blood pressure control over 48 hours with EVD discontinuation on sixth day. Second patient was a 44 year old African American man who presented with subacute headaches, lightheadedness, gait instability and bilateral papilledema. BP on admission was 242/118 mm Hg. Head CT W/O contrast showed hypodensity of bilateral cerebellar hemispheres, effacement of the basilar and peri-pontine cisterns with obstruction of the fourth ventricle. EVD was placed with opening pressure of 30 cm H20. MRI brain revealed hyperintensities diffusely involving the cerebellum, right anterior aspect of the pons, midbrain and midline of the medulla. There was tonsillar herniation and obstruction of the fourth ventricle. Blood pressure control over 48 hours and CSF drainage resulted in resolution of symptoms with discontinuation of EVD in 7 days.

RBHE is a rare reversible cause of obstructive hydrocephalus. Early detection, aggressive blood pressure management with temporary CSF diversion resulted in good outcomes.

Coma is a marker of poor outcome in ICU patients. Acute encephalitis presenting with prolonged seizures is a common etiology for coma. With aggressive neuro-ICU care, recovery is possible; however, no reliable predictor of return of consciousness (ROC) has been established. Notably, physiologic sleep has been associated with favorable outcome in certain acute brain injuries. We evaluated the relationship between return of physiologic sleep patterns on EEG and ROC and functional outcome following encephalitis.

Case review of two comatose patients with encephalitis and refractory status epilepticus and literature review.

Both patients had encephalitis of unknown etiology and presented with coma and refractory status epilepticus. Patient one, a 33 year old male, was treated with antivirals and 9 antiepileptic drugs (AEDs) including high dose phenobarbital, ketamine, empiric steroids, plasma exchange and eventually the ketogenic diet. His seizures resolved by day 38, EEG demonstrated vertex sharp waves and sleep spindles on day 39, and ROC occurred 6 days later on day 45. By discharge to rehab, he was awake and oriented with intact language, was feeding himself, and had normal motor power throughout. Patient two, a 23-year-old male, was treated with antivirals and 12 AEDs. His seizures resolved on day 81 and cEEg demonstrated sleep-wake cycling and rudimentary sleep spindles on day 83. AEDs were weaned and he followed commands 4 days later. By discharge he was awake and oriented with intact language, and had no focal motor deficits.

The return of physiologic sleep patterns on EEG in two patients with severe brain injury from encephalitis appeared to herald the recovery of consciousness and favorable functional outcome. This electrographic finding may inform treatment and prognosis. These findings need to be confirmed in a larger prospective study.

Therapeutic hypothermia following cardiac arrest has been shown to favorably impact survival and neurologic outcome. The risk of cooling a patient with scleroderma and Reynaud's phenomenon would appear to be significant.

A 45 year old female with end stage renal disease, CAD, scleroderma, and Reynaud's phenomenon was undergoing her daily dialysis when she arrested. Initial EKG showed ventricular fibrillation. CPR/ACLS was initiated with ROSC in 20 minutes. She did not immediately regain consciousness. Patient was cooled to 33 degrees and maintained at that temperature for 24 hours per protocol. Surface cooling was necessitated because of access issues. She was counterwarmed from the initiation of hypothermia using active warming blankets, Rooke boots, and Rooke mitts. Norepinephrine drip was needed for blood pressure support in the first 24 hours. Extremities were closely monitored for signs of vascular compromise during cooling and re-warming phase. Rewarming was carried out at 0.2 degrees per hour. Patient was following commands and extubated by day 4 post-arrest with neither neurologic deficit nor ischemia of extremities or digits.

Patients with scleroderma and Reynaud's phenomenon would appear to be at very high risk for ischemic complications of therapeutic hypothermia. There is a previous case report that details a scleroderma/Reynaud's patient who suffered digit ischemia after cooling for cardiac arrest. In our rather frail 50kg patient, we chose to aggressively counter-warm and protect her extremities from the outset of cooling because she had previously lost a digit to ischemia. Intravascular cooling would have been our first choice in this patient, but was not an option due to access graft in one groin and a previous vascular reconstruction of the other. The risk of therapeutic hypothermia in a patient with Reynaud's might discourage the provider from considering its use in such seemingly high risk paitients. We have demonstrated that it can be successfully accomplished in a patient who has previously suffered ischemic complications of Reynaud's.

In few countries death is declared after brainstem death, but not in the USA. In this country, death is determined when there is irreversible cessation of all functions of the entire brain, including the brainstem. We present a patient, who for 6 days met clinical criteria for brainstem death, but subsequently became brain dead (BD).

A 41 year-old man was admitted for subarachnoid hemorrhage (admission Glasgow Coma Scale 15, Hunt&Hess 1, Fisher 4). An angiogram showed an irregular right posterior inferior cerebellar artery (PICA) segment and he was treated medically in the NICU for 12 days. He then underwent a suboccipital craniotomy with ligation and bypass of the PICA segment with an arterial graft. Post-operatively, he remained in the NICU for several days, eventually requiring tracheostomy and percutaneous gastrostomy. One month after admission he developed an acute neurological decline and new posterior fossa hemorrhage. Emergent evacuation of the hematoma was unsuccessful as profound operative bleeding was encountered. A repeat angiogram showed occlusion of the graft and significant blood extravasation over the left vertebral artery (VA), necessitating multiple coils to occlude the left VA. He remained comatose without any spontaneous or reflexive movements for 6 additional days in the NICU, but with normal blood flow in the anterior circulation by Transcranial Dopplers (TCD) and SPECT (on the 5 th day). On the 6th day, as both TCD and SPECT showed absence of intracranial blood flow, the patient was pronounced BD and became an organ donor.

This case illustrates the challenges encountered with brainstem death and the disconnect between clinical and paraclinical results, when BD confirmation is attempted. It also demonstrates that with time these patients may eventually become BD, probably due to obstructive hydrocephalus, although the natural progression of this entity is relatively unknown and possibly not uniform.

Since the 1980s the use of computed tomography (CT) has increased more than 20 fold; its widespread availability and rapid acquisition time make it a valuable diagnostic tool. Unlike magnetic resonance imaging (MRI), CT exposes patients to ionizing radiation. The Linear Hypothesis theory assumes all radiation exposure, irrespective of the dose, is detrimental to living tissue with no safety threshold. While controversial, this has led to increased radiation precaution and exposure limitation guidelines. Whereas whole brain CT perfusion spreads radiation over a large area, two-slice CT perfusion (CTP) focuses radiation at the level of the basal ganglia. We report the development of focal 'stripe' alopecia from exposure to CT angiography (CTA) and focused beam 2-slice CT perfusion, a transient complication of focal beam radiation.

We reviewed all cases of CT induced alopecia from 2008 to 2013 in the neurologic intensive care unit at Mayo Clinic Florida. We describe 4 patients (2 male; 2 female) with mean age 35 (range 27-43). The underlying diagnoses were subarachnoid hemorrhage (3) and carotid artery dissection (1).

Each patient had more than 2 studies per week (CTA or CTA/CTP) during hospitalization. The radiation dose was estimated to be 2-3 Gy; stripe alopecia occurred in 4 out of 4 patients (100%) but was transient with hair growth returning in approximately 1 month. To date, 0/4 patients have been diagnosed with any secondary hematologic or intracranial malignancy.

Critically ill patients with carotid dissection or vasospasm from subarachnoid hemorrhage who undergo 2 or more CTA or CTP studies per week with a 2-slice focused beam method may develop transient 'stripe' alopecia. Decreasing the frequency of focused beam CT or using whole brain perfusion may help avoid this radiation complication. Exposure to ionizing radiation should be limited by following the As Low As Reasonably Achievable (ALARA) principle.

Tetanus results from the toxin of Clostridium tetani producing a systemic disinhibition of neuromuscular transmission. The tetanus toxin (tetanospasmin) travels in a retrograde fashion from peripheral nerve terminals to the spinal cord and brainstem motor inhibitory neurons, resulting in diminished release of inhibitory neurotransmitters, such as GABA and glycine, and subsequent tetanic stimulation of skeletal muscle. Tetanus is relatively rare in the developed world due to aggressive vaccination efforts against the tetanus toxin. However, in those who cannot tolerate the vaccine, tetanus remains a very real threat.

Case report with literature review.

We describe a 79 year-old woman without vaccination against tetanus toxoid due to anaphylactic allergy to tetanus toxoid vaccine who presented to the emergency room with inability to swallow or open her mouth two weeks after tearing the skin on her right arm on a metal stool. Her exam revealed trismus and a "spatula reflex," in addition to right upper extremity rigidity and an erythematous, draining wound to the right arm. Human tetanus immunoglobulin was administered, as was botulinum toxin A injection (100 mouse units total) to the masseters. Mild improvement in trismus was noted 3 days later. Diphtheroid bacilli were isolated from the wound on culture. Electrophysiologic study showed an absent masseter inhibitory reflex with the jaw jerk, supporting an electrophysiological diagnosis of tetanus. The patient was admitted to the intensive care unit for intubation due to neuromuscular respiratory failure. Ultimately, the patient died from distributive shock and ARDS from ventilator-associated pneumonia.

Tetanus remains a rare diagnosis in the developed world due to aggressive vaccination efforts, but remains a risk for unvaccinated patients, and can have a complicated ICU management. Botulinum toxin A, a toxin derived from C. botulinium, ironically may offer a temporizing treatment for severe trismus and dystonia due to tetanospasmin.

The mediastinum is a rare site for occurrence of teratomas, as most documented cases are ovarian in origin. In this particular instance, the patient did not present with the most common signs of shortness of breath or retrosternal pain or fullness, but instead with an acute psychotic episode and refractory non-convulsive status epilepticus. Unique presentations such as this one often force healthcare providers to think "outside the box" when determining the cause of non-convulsive status. In late 2009, a previously healthy 19 year old female college student began having emotional labiality which quickly progressed to uncharacteristic hostility, religious preoccupation, and visual hallucinations. Shortly thereafter, she began seizing despite the fact that her head CT, MRI, and LP all lacked acute abnormalities. She was intubated and transferred to the neuro intensive care unit at our tertiary care center. Upon arrival patient was noted to have nystagmus, and rhythmic facial movements, as well as an EEG confirming non-convulsive status. The status epilepticus was refractory to a variety of treatments including; benzodiazepines, multiple anti-epileptic drugs (phenytoin, levetiracetam, oxcarbazepine, valproate, and phenobarbital), repeated drug induced coma, resection of a mature mediastinal teratoma, intravenous steroids, rituximab, and plasmapheresis. Her seizures were eventually controlled using felbamate. Once the neoplasm was removed, the episodes of psychosis resolved suggesting they were likely secondary to the anti-NMDA antibodies. Besides status, her critical care course was complicated by limbic encephalitis, acute respiratory failure requiring tracheostomy, and bilateral pneumothoraces. After a prolonged hospital course, she was eventually discharged with minimal functional deficits, and able to return back to college full-time.

This case represents a unique presentation of mediastinal teratoma and non-convulsive status epilepticus. When considering the complex differential diagnosis of acute psychosis, critical care providers may need to consider psychiatric symptoms as potential early manifestations of non-cranial paraneoplastic syndrome.

Language disturbance is common symptom encountered by clinical neurologists. It is often caused by an acute ischemic stroke involving the left middle cerebral artery. Whereas speech arrest is not uncommon during epileptic seizures, isolated language disturbance is extremely rare. Astrocytoma, meningioma, and idiopathic cause have been reported as etiologies of prolonged ictal aphasia. Here we report three cases with other etiologies presenting with prolonged ictal aphasia.

A-74-year old woman presented with motor aphasia for 2 days. No accompanied symptoms, such as convulsive movement, hemiparesis, were noted. No acute infarct lesion was detected by brain MRI. But gadolinium enhanced MRI showed large multiple linear enhancing lesions, suggesting vascular malformation. Left frontotemporal sharp waves were frequently appeared on electroencephlogram(EEG). The symptom and abnormal discharges were disappeared after injection of intravenous lorazepam. A-78-year old man admitted to our hospital because of bacterial meningitis. During treatment, motor aphasia had been developed. Brain MRI showed an intracranial hemorrhage(ICH) in left temporal lobe. But motor aphasia could not explained by lesion of ICH. Diffusion weighted MRI showed a high signal intensity on left frontal cortical vein and EEG showed periodic lateralized epileptiform discharges(PLEDs) in left hemisphere. PLEDs were disappeared and his symptom was improved by using intravenous lorazepam. A-42-year old man presented with sensory aphasia for 1 days. No previous history of seizure, trauma, hemiparesis was noted. Brain MRI showed tubular calcification along cortical portion at left temporal lesion and linear enhancing lesion in adjacent pachymeninges, suggesting meningioangiomatosis. Left temporal spikes on EEG was changed to left fronto-temporal slow waves after induction of intravenous lorazepam. Also his symptom was improved simultaneously.

Ictal aphasia may occur as an isolated manifestation without other evidence of seizure activity.

An electroencephalography with intravenous benzodiazepine should be performed in all unexplained prolonged aphasia to diagnose epileptic seizure involving the language cortex.

Trismus is a motor disturbance of the trigeminal nerve , especially spasm of the masticatory muscles, resulting in difficulty in opening the mouth. It has a number of potential causes which range from the simple and non-progressive to those that are potentially life-threatening. In this paper we report for the first time the occurrence of trismus in the post ictal state.

This was a case study which was conducted at King Khalid Hospital.

After a series of generalized convulsive seizures the patient developed a sustained trismus in the postictal phase, lasting for three days.

A number of physiological and metabolic factors have been implicated in the termination of seizure activity and transition to post-ictal state by creating inhibitory signals. In our patient the trismus may have signified a partial failure of the aforementioned inhibitory mechanisms at the mesencephalic-pontine level , causing a disruption of projections of the mesencephalic trigeminal nucleus to the pontine nucleus, resulting in a state of hypertonicity in the latter.

Spinal artery aneurysm is a rare cause of subarachnoid hemorrhage.

We describe a case that presented to our medical center.

A seventy-five year old woman with Parkinson's disease presented with acute chest pain, followed immediately by severe headache. On examination, her blood pressure was 116/74. She had profound meningismus on passive mobilization of the neck, and required vigorous stimulation to maintain wakefulness. A CT scan of her head demonstrated perimesencephalic subarachnoid hemorrhage, but a CT angiogram with contrast did not show any intracranial vascular abnormalities. To evaluate her chest pain, a body CT with contrast was performed, and showed a linear enhancing lesion within the spinal cord at levels T10-L1. A conventional spinal angiogram revealed a 5mm, partially thrombosed, fusiform aneurysm at the level of L1, arising from a descending ramus of the anterior spinal artery. MRI of the total spine confirmed the presence of diffuse blood products throughout the subarachnoid space anterior to the cord, supporting rupture of this aneurysm and upward extension into the perimesencephalic space as the cause of her headache and chest pain. The aneurysm was surgically accessed via a T11-T12 laminectomy and clipped. A follow up spinal angiogram showed no residual filling of the aneurysm. The patient's symptoms improved post-operatively and she was discharged to an acute rehabilitation facility.

Subarachnoid hemorrhage from rupture of a spinal artery aneurysm can present with a variety of symptoms including headache, back pain and chest pain. A conventional spinal angiogram should be considered in the diagnostic evaluation of idiopathic subarachnoid hemorrhage.

Traditional treatment for elevated intracranial hypertension is well known and includes osmotic therapy, hyperventilation, and sedation. The consensus is that refractory cases can be treated with decompressive craniectomy, barbiturate coma, or most recently, therapeutic hypothermia. In cases that remain refractory to craniectomies, a lobectomy is sometimes performed. This procedure can result in resection of viable tissue potentially limiting functional outcomes. To date, there are no trials that suggest the use of hypothermia after craniectomy. We present a case of a high grade aSAH with refractory intracranial hypertension after bifrontal craniectomy who was then successfully treated with therapeutic hypothermia.

HR is a 29 y/o Hispanic male who presented with a grade V aSAH. On day 8, he was successfully treated for vasospasm with angioplasty. Later he developed refractory intracranial hypertension for which therapeutic hypothermia was initiated and a decompressive bifrontal craniectomy was performed. Twenty-four hours after surgery, re-warming was initiated. At 36°C, the ICP increased to 27 mmHg and hypothermia was resumed. The patient required hypothermia to control ICP for 9 additional days and was eventually re-warmed on day 18.

HR was discharged to long term care six weeks after presentation and completed his recovery at a rehabilitation center. 10 months after his initial event, HR has made a full recovery and hopes to return to work by the end of the year.

Some facilities choose lobectomy as a salvage therapy to treat life-threatening refractory intracranial hypertension. In this case, we were able to overt the possibility of removing viable brain tissue in an effort to control intracranial hypertension through the use of therapeutic hypothermia as an adjunct to bifrontal craniectomy which allowed this patient to return to his premorbid state.

The introduction of minocycline-rifampin coated external ventricular drains (EVDs)into clinical practice was designed to reduce the incidence of Gram-positive nosocomial ventriculitis. In our institution, the rate of nosocomial ventriculitis after the introduction of antibiotic-coated EVDs in the past 7 months was 4.1 infections per 1,000 EVD-days (or, 2 infections/64 insertions; 3.1%); one of which was by Candida albicans. The risk of fungal ventriculitis in immunocompetent individuals with non-antibiotic coated EVDs is traditionally quite low. However, with suppression of bacterial infections due to EVD antibiotics, the proportion of fungal ventriculitides may rise. We report a case of fungal ventriculitis in a patient with a minocycline-rifampin coated EVD.

A 59 year old male presented with a Hunt-Hess 2, Fisher 3 subarachnoid hemorrhage secondary to a ruptured basilar tip aneurysm. A minocycline-rifampin coated EVD was placed on admission using standard aseptic technique and 1 dose of peri-insertional cefazolin. The patient had a medically complicated post-embolization course (hydrocephalus, cerebral salt wasting, ventilator-associated pneumonia, and deep venous thrombosis), prolonging the need for external ventricular drainage. On day 16 post-EVD placement, the patient became febrile, with negative urine, blood, and sputum cultures; but with cerebrospinal fluid (CSF) studies consistent with ventriculitis. CSF Gram stain and cultures confirmed Candida albicans. Intravenous fluconazole was initiated and the EVD was exchanged. CSF cultures became negative on hospital day 19.

This is the first and only published report of Candida ventriculitis following placement of a minocycline-rifampin coated EVD. We suspect that less traditional ventricular infections due to pathogens like fungus, Gram-negative bacteria, and resistant Gram-positive bacteria will require a higher index of suspicion with suppression of Gram-positive bacteria due to antibiotic-coated EVDs. Rates of fungal, Gram-negative, and resistant forms of Gram-positive ventriculitides should be monitored in institutions using antibiotic-coated EVDs.

Mild therapeutic hypothermia (TH) is recommended for neuroprotection in those patients who remain unresponsive following resuscitation from cardiac arrest. However, the optimal management of other aspects of care, including hemodynamic status, remains uncertain. Hemodynamic fluctuations are frequent during the phases of TH, and the adequacy of cerebral oxygenation during these fluctuations is often unclear. Continuous measurement of jugular venous oxygen saturation (S J VO 2 ) may quantitatively assess cerebral oxygen delivery and extraction, though there is minimal literature describing the use of this advanced neuromonitor in post-arrest patients, particularly in regards to guidance of hemodynamic goals during TH.

Design -Case report.

A 70 year-old male with an extensive cardiac history was brought to our emergency department following an out-ofhospital cardiac arrest due to ventricular fibrillation. Post-resuscitation, he was unresponsive and met criteria for the initiation of therapeutic hypothermia (TH). A post-arrest echocardiogram revealed evidence of severe left ventricular dysfunction. To assess the adequacy of cerebral perfusion during TH, continuous jugular venous oximetry was performed. While hypothermic, the patient became bradycardic with a heart rate in the 30s. Simultaneously, the jugular venous oxygen saturation (S J VO 2 ) decreased to 37%, indicative of global cerebral ischemia even though the patient's systolic blood pressure was greater than 90 mm Hg. A transvenous pacing wire was introduced and the patient was ventricularly paced at a rate of 80. Shortly thereafter, the patient's S J VO 2 normalized. The patient recovered significant neurologic function throughout the remainder of his hospital stay and ultimately had a highly favorable short-term functional outcome.

This case report is an example of the use of advanced neuromonitoring techniques to 1) identify episodes of cerebral hypoperfusion in cardiac arrest survivors, and 2) ascertain the restoration of adequate perfusion during therapeutic maneuvers. We believe continuous jugular venous oximetry may be a useful adjunct to the neuroprotective efforts of therapeutic hypothermia.

Malignant cerebral infarction has a high risk of fatal brain edema and increased intracranial pressure with cerebral herniation causing death. Currently, therapeutic hypothermia has been suggested as a potential treatment in large hemispheric infarction.However, in them receiving therapeutic hypothermia, a rebound cerebral edema during rewarming phase was reported as the major cause of death and definite treatment is uncertain.

A 66-year-old male patient presented with the right hemiplegia and global aphasia was transferred to emergency room after intravenous tissue plasminogen activator (alteplase) was administered at local hospital. Outside CT and CT angiography revealed malignant cerebral infarction in the whole territory of middle cerebral artery with the occlusion of the proximal internal carotid artery and follow-up MRI and MRA at our hospital showed no change. Being refused decompressive hemicraniectomy, he received the therapeutic hypothermia with mannitol. After finding the decline of the cerebral edema on CT, rewarming was decided and the rate was 0.5 /hr. The duration of therapeutic hypothermia was about 6 days (160 hours). As finishing the rewarming phase, mentality was suddenly decreased and dilated left pupil. Follow-up CT revealed that midline shifting was more aggravated. We decided on repeated hypothermia for the rebound edema and successfully controlled without serious adverse effects. The second-rate of rewarming was 0.1 /hr. Total duration of repeated hypothermia was about 4 days (100 hours).

Therefore, we reported our experience with repeated hypothermia for rebound cerebral edema following therapeutic hypothermia in malignant cerebral infarction. Repeated hypothermia may be one of the therapeutic options to control rebound edema.

Financial Support: None

Title Presenting Author 281 Recovery of Consciousness in Comatose Patients with Acute Encephalitis and Super Refractory Status Epilepticus: Impact of EEG Sleep Intrusions Benjamin J. Barnes 282 Therapeutic Hypothermia Without Ischemic Complications Following Cardiac Arrest in a Patient with Scleroderma and Reynau's Phenomenon Charles J. Miller 283 Brainstem death is not equivalent but leads to brain death Panayiotis N. Varelas 284 Transient "Stripe

Acute Psychosis as the initial presentation of a 19yo with Medistinal Teratoma Justin Calabrace 287 Report of three cases presenting with prolonged ictal aphasia Yun-Ju Choi 288 Protracted Post-Ictal Trismus Muhammad U. Khan 289 The value of arterial spin labeling cerebral blood flow imaging in the diagnosis of seizures

Subarachnoid Hemorrhage from Ruptured Spinal Artery Aneurysm Sheila Chan 291 Lobectomy versus Therapeutic Hypothermia for Refractory Intracranial Hypertension After Bifrontal Craniectomy? That is the Question Patty Gessner 292 A Case of Candida Ventriculitis in a Patient with an Antibiotic-Impregnated External Ventricular Drain. Do Antibiotic-Impregnated External Ventricular Drains Confer a New Selection of Micro-Organisms?

Hoshide 293 "Normal Pressure Herniation" with Progressive Neurological Deterioration after Poor Grade Aneurysmal Subarachnoid Hemorrhage and Global Cerebral Edema

Jugular Venous Oxygen Monitoring to Improve Cerebral Oxygen Delivery in a Patient Undergoing Therapeutic Hypothermia after Cardiac Arrest

Edwards 295 Repeated Hypothermia for Rebound Cerebral Edema after Therapeutic Hypothermia in Malignant Cerebral Infarction Jeong-Ho Hong 296 Intracerebral microdialysis parameter (especially glycerol) may indicate the severity of the brain damage in patients with post-cardiac arrest syndrome with

Yale New Haven Hospital/Neuroscience ICU

Acknowledgments:NeMoDevices provided funding for this clinical trial. Acknowledgments:NeMoDevices provided funding for this clinical trial.

Shirin Jalini 1 

Measurement of brain oxygen tension (PbtO 2 ) is clinically useful in traumatic brain injury (TBI). We describe changes in PbtO 2 during therapeutic hypothermia (TH) and examine PbtO 2 's association with outcome.

A post-hoc analysis of the National Acute Brain Injury Study:Hypothermia (NABISH:II) trial was performed. PbtO 2 monitors were inserted within 8hrs and values recorded hourly. Probes were placed into normal brain or opposite side of hematoma/craniectomy. Treatment of PbtO 2 was at the clinical team's discretion. We examined PbtO 2 during TH (10-48hrs), prior to rewarming.

97 patients were enrolled in NABISH:II. After exclusion for malfunctioning probes and/or <8hrs of data, 83 patients were included for analysis [37 normothermia (NT), 46 hypothermia (HT)]. General demographics, including age, injury and GCS were not different between NT and HT. No difference in death at 30 days was found (5(13.5%), 9(20%); p=.46). Using thresholds of 6, 10, 15 and 20 mmHg, no differences in risk of brain tissue hypoxia were observed between NT versus HT or between survivors versus non-survivors. Mean PbtO 2 was lower in HT (20.2mmHg, 26.3mmHg; p=0.006). ICP and CPP were not different between treatment groups. A trend for higher MAPs were seen in HT. To account for repeated observations per patient, a mixed-effects model was used. While controlling for baseline demographics and CPP, PbtO 2 was 6.4mmHg lower in HT (p<0.01). Mean PbtO 2 was no different between non-survivors and survivors (20.8mmHg, 23.4mmHg;p=0.38) ICP was higher and CPP was lower in the non-survivors (22.1mmHg, 13.5mmHg;p<0.01) (76.6mmHg, 85.4mmHg p<0.01), respectively. MAP was no different. A mixed-effects model showed no difference in PbtO 2 between non-survivors and survivors (p=0.5).

Hypothermia after TBI was associated with lower PbtO 2 . No differences were seen in risk of tissue hypoxemia using defined thresholds. No association between PbtO 2 and mortality was seen. 

Epileptic seizure leads to a decrease in brain tissue oxygen tension even with an increase in cerebral blood flow (CBF). Arterial Spin Labeling (ASL) perfusion is a non-invasive magnetic resonance imaging technique without need of contrast agent, which is especially useful in patients in the intensive care unit (ICU) because they may have deteriorated renal function. Therefore, we hypothesized that hyperperfusion pattern on ASL map may be useful in identifying ictal hyperperfusion zone in the ICU.

From January 2012 to June 2013, 7 patients with documented seizure (3 had convulsive seizures, and 4 had nonconvulsive status epilepticus) on electroencephalogram (EEG) underwent ASL perfusion MRI within 15 minutes after the cessation of seizures. The absolute CSF value was measured and correlated with epileptiform discharge on EEG

All with convulsive seizure had higher CBF in the corresponding cortex (55.4 ± 11.3 ml/100g/min) compared to the contralateral mirror area (35.6 ± 8.6) ( P<0.01) on ASL map. In patients with non-convulsive seizure, CBF was elevated in the bilateral thalami (52.1 ± 13.1 ml/100g/min) and superior colliculus (48.3± 9.2 ml/100g/min).

Our preliminary case series showed that ASL perfusion might be a useful tool in identifying seizure related hyperperfusion pattern in patients in the ICU. The meaning of different perfusion pattern in convulsive and non-convulsive seizure requires further study.

Tamra I. Ranasinghe, Rabih Tawk, Ramon Navarro, William D. Freeman Mayo Clinic, Jacksonville, FL, USA

Herniation of brain tissue typically occurs from a region of relative higher tissue or intracranial pressure (ICP) or with globally elevated ICP (>20mmHg). However, herniation can occur with normal (<20mmHg) ICP values. We describe a case of normal pressure herniation ("NPH") from global cerebral edema with normal ICP values.Methods/Result: Case report. We describe a 46 year old male who presented with coma (GCS =7) after a severe (modified Fisher 4) aneurysmal subarachnoid hemorrhage from a ruptured anterior communicating artery (ACOMM) aneurysm with bilateral frontal intraparenchymal hemorrhages (IPH). The aneurysm was coiled and an external ventricular drain placed to monitor ICP. Several hours later, he underwent bifrontal craniectomy due to refractory elevated ICP (>20mmHg) despite mannitol, hypertonic saline, and had the frontal IPH's evacuated but the bone flap replaced. The CT scan demonstrated global cerebral edema (GCE) with diffuse effacement of sulci and loss of space around brainstem. On postoperative day 5, the patient had progressive loss of pupil reactivity (3mm reactive OU to 7mm unreactive OU) with downward brainstem herniation on MRI despite ICP's ranging from 8-18mmHg. The patient underwent redo bifrontal craniectomy with bone flap removal. The patient remained comatose and repeat MRI demonstrated progressive downward herniation of the midbrain and pons into the posterior fossa despite ICP values being <15mmHg. The family eventually withdrew care due to vegetative state.

"Normal Pressure Herniation" can occur with ICP values <20mmHg due to intracranial compartment gradients in pressure or edematous tissue. Craniectomy was ineffective in preventing mechanical tissue shift of forebrain into the posterior fossa and perhaps worsened this due to atmospheric pressure influences (760mmHg) on frontal brain regions. Reported cases of NPH and even "low pressure herniation" are rare and require further study into salvage therapies can prevent irreversible brain injury and brain death. 

Therapeutic hypothermia has proven to improve outcome in patients with post-cardiac arrest brain injury (PCBI). This study evaluates the neurochemical changes indicating cerebral membrane degradation (glycerol) and excitotoxic process (glutamate) in patients with PCBI with therapeutic hypothermia (TH).

Ten adult comatose patients (25-76 years old, 6 male, 4 female) with PCBIwere enrolled. Neuromonitoring was performed by the consist of cerebral microdialysis (MD) and intracranial pressure (ICP) measurements. These probes were placed in right frontal cortex. MD were taken every hours and the concentrations of MD parameters were analyzed enzymephotometrically.

Initial values of glycerol (396+-438 microM, mean +-SD) and glutamate (39+-22 microM) are high, which are measured within 4 to 10 hours after return of spontaneous circulation (ROSC). Glycerol in two patients with high ICP and poor outcome shows extreme increases (over 1000 microM). These increases of glycerol and glutamate normalize within a couple of days, during the induction and maintenance phase of TH. During re-warming phase (It takes 24 to 72 hours), reincreases of MD parameters were not clearly evident. Glycerol concentrations rose immediately and transiently after every intravenous glycerol (osmotic diuretic) administration in three cases of diuretic usage. Glycerol increases from 192+-242 to 500+-390 microM (max. 1600) even after 48 hours from ROSC in one case with poor outcome.

Intracerebral MD may indicate the severity of the brain damage. MD parameters are informative during each phase of TH.Financial Support: None