key: cord-302997-39o08tt1 authors: Ceruti, S.; Roncador, M.; Gie, O.; Bona, G.; Iattoni, M.; Biggiogero, M.; Maida, P. A.; COVID-19 Clinical Management Team,; Garzoni, C.; Mauri, R. title: Reduced mortality and shorten ICU stay in SARS-COV-2 pneumonia: a low PEEP strategy date: 2020-05-08 journal: nan DOI: 10.1101/2020.05.03.20089318 sha: doc_id: 302997 cord_uid: 39o08tt1 Background Intensive Care Unit (ICU) management of COVID-19 patients with severe hypoxemia is associated with high mortality. We implemented a "care map", as a standardized multidisciplinary approach to improve patients monitoring using: uniform patient selection for ICU admission, a low-PEEP strategy and a pharmacologic strategic thromboembolism management. Methods A standardized protocol for managing COVID-19 patients and ICU admissions was implemented through accurate Early Warning Score (EWS) monitoring and thromboembolism prophylaxis at hospital admission. Dyspnea, mental confusion or SpO2 less than 85% were criteria for ICU admission. Ventilation approach employed low PEEP values (about 10 cmH2O in presence of lung compliance > 40 mL/cmH2O) and FiO2 as needed. In presence of lower lung compliance (< 40 mL/cmH2O) PEEP value was increased to about 14 cmH2O. Results From March 16th to April 12nd 2020, 41 COVID-19 patients were admitted to our ICU from a total of 310 patients. 83% (34) of them needed mechanical ventilation. The ventilation approach chosen employed low PEEP value based on BMI (PEEP 11+/- 3.8 (10-12) cmH2O if BMI < 30 Kg/m2; PEEP 15+/- 3.26 (12-18) cmH2O if BMI >30 Kg/m2). To date, ten patients (24%) died, four (9.7%) received mechanical ventilation, two were transferred to another hospital and 25 (60.9%) were discharged from ICU after a median of nine days. Discussion A multimodal approach for COVID-19 patients is mandatory. The knowledge of this multi-organ disease is growing rapidly, requiring improvements in the standard of care. Our approach implements an accurate pre-ICU monitoring and strict selection for ICU admission, and allows to reduce mechanical ventilation, ICU stay and mortality. Funding No funding has been required. 4 echocardiography (to establish the global cardiac function before any pronation). A thoracic CT-scan was considered available if it has been performed during the stay in the Internal Medicine Department during the last 24 hours before the admission in the ICU. Indication for ICU admission and oro-tracheal intubation (OTI) was routinely established by the intensive care specialist or senior anesthesiologist on duty, according to the 'care map' based on standardized criteria selection, low PEEP strategy and pharmacologic antithrombotic management. Requests of counseling for ICU admission came from the Department of Internal Medicine and from the Emergency Department (ED). With the aim of quickly identifying the worsening of clinical conditions 11,12 , all consultations were recorded by reporting patient's symptoms, SpO 2 , blood gas analysis values (if available) and clinical decision for admission or not to the ICU. Patients presenting partial respiratory failure combining peripheral saturation (SpO 2 ) lower than 85% and dyspnea (or mental confusion), or patients with dyspnea (or mental confusion) alone, were eligible to be admitted in ICU. Exclusion criteria were the will of the patient not to be intubated, cardiocirculatory arrest following hypoxia, metastatic oncological disease, end-stage neurodegenerative disease, severe and irreversible chronic disease (heart failure NYHA IV, COPD GOLD D, liver cirrhosis Child-Pugh > 8, severe dementia) 10 . With the aim to avoid a misleading interpretation in ICU mortality, we decided to perform an extra evaluation on patients excluded from ICU, to ensure about their survival status. After endotracheal intubation, we initially provided low PEEP-value strategy based on BMI (PEEP 10 cmH 2 Intensive supportive care was managed according to the evolution of the inflammatory parameters (CRP, CK, LDH and ferritin) and the stability of the P/F-ratio after each supination during the following days. In case of a favorable evolution, sedation was reduced to RASS -3/-2 switching on Propofol. By improvement of blood oxygen levels, we proceeded to reduce FiO 2 up to 35-40% FiO 2 values without reducing PEEP. Once all the clinical and biological inflammatory parameters were constantly reduced for almost three consecutive days, patients were gently weaned from PEEP by keeping a PaO 2 > 60 mmHg (8 kPa) . The choice of removing the endotracheal tube was made by the doctor in charge according to usual standard of care. Deep vein thrombosis, PE, Ventilator-Associated-Pneumonia (VAP) and Acute Kidney Injury (AKI) have been the main complications arose in patients admitted to ICU: DVTs and PEs were defined as suspected with an increase in serum D-dimer values over 1,500 ng/ml, while they were considered as confirmed by ultrasound or CT-scan positive finding, defined according to current clinical standards. VAPs were identified according to usual standard care 16 with an increase in secretions, in their quality and quantity, requiring an increase in the FiO 2 administration 17 . Each case of AKI RIFLE F requesting CVVHDF has also been reported. All complications, administered drugs and adverse events occurring during the stay in intensive care, were registered and reported in the electronic medical record. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 8, 2020. . Descriptive statistics were used to summarize the clinical data collected. No statistical sample size calculation was performed. We present continuous measurements as mean (min-max, SD) otherwise as median (IQR) if they are normally distributed. Categorical variables were reported as counts and percentages. Test statistics and survival analysis were performed with R v.3.6.1 and the Kaplan-Meier estimator from CRAN "surv" package. Data was subsequently compared with a similar in number and follow-up cohort published by Bhatraju et al 18 - complete patient data was retrieved from the supplementary appendix. This study has been notified to the Ethics Committees of Canton Ticino. According to the local Federal rules, it has been approved as a clinical data collection case series. No funding has been required. The corresponding author confirms that he had full access to all the data in the study and he had final responsibility for the decision to submit for publication. 310 patients with COVID-19 symptoms presented to our Clinic. According to exclusion criteria, 54 of them were not admitted to the ICU as they had a "do not resuscitate" order (DNR) in place before hospital admission Table 1 . Patients had a mean age of 64±16·5 years; most of them were men, often burdened by one or more chronic medical conditions (Table 1) . At ICU admission, most patients showed hemodynamic stability. A chest CT-scan was obtained in 23 (56%) patients; all of them showed bilateral ground glass opacities and four of them showed consolidations in addition (Table 1) . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. . https://doi.org/10.1101/2020.05.03.20089318 doi: medRxiv preprint At admission PaO 2 :FiO 2 ratios had a median of 87 (54 -133); with a median FiO 2 of 72% (60% -100%) at the ICU admission. Thirty-two (78%) patients received invasive MV, the others were treated with High-Flow Nasal Cannula (Table 2) . Globally, mean PEEP for patients with BMI < 30 Kg/m 2 was 11 cmH 2 O (10-12, SD 3·80), while mean PEEP for patients with BMI > 30 Kg/m 2 was 15 cmH 2 O (12-18, SD 3·26). After the onset of MV, the median FiO 2 improved around 70% (60% -85%), with first PaO 2 :FiO 2 ratio with a median of 147 (101 -233) ( Table 2) . Thirty-one (75·6%) patients were placed in a prone position (with an average number of pronations of four). In these patients, PaO 2 :FiO 2 ratio progressively improved during next days, with a median value of 100 (67·5 -153) during the first day, 142·5 (97·6 -232·8) during the second day and 167 (113 -230) during the third day (Table 3 ). No patient presented any contraindication to be treated with parenteral anticoagulation; 15 (36%) patients were simply treated through prophylaxis, while 26 (63%) patients were managed by full therapeutic anticoagulation (12 -46% -with Unfractioned Heparin, 14 -54% -with LMWH). No patient presented any bleeding complication, nor clinical sign requiring anticoagulation reduction or removal. The median length of ICU stay was nine days (4 -12·5); the median duration of MV was seven days (3 -10) (Table 3) . On 12 th April, of the 41 patients, ten (24·3%) has died, four (9·7%) are still in the ICU receiving MV (two endotracheal tube, two tracheostomy), two patients were transferred to another hospital and 25 (60.9%) have been discharged from the ICU ( Figure 2 ) in good medical condition with no additional death in the following 10 days. At 7 day from ICU discharge, 39 patients (95%) presented a Karnofsky performance status of more than 80. No patients have been reintubated within or after 72 hours. The Intensivist Consultant performed clinical counseling of 48 patients admitted to the Internal Medicine Department, ED or from other hospital. The mean age was 65±25 years (38 -82), presenting SpO 2 median of 90% (88 -94), PaO 2 median of 63·2 mmHg (51·35 -76·55), a pCO 2 median of 35·2 mmHg (31·9 -39·1) and a median value of Hb of 13·9 g/L (12·3 -15·4). Twenty-six of these patients moved to our ICU and five of them . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. . https://doi.org/10.1101/2020.05.03.20089318 doi: medRxiv preprint had already been intubated in other hospitals. Intriguingly, the remaining 22 patients consulted but not admitted improved their conditions from an initial median value of SpO 2 of 87% (84-91%) without being highly symptomatic nor requesting admission in ICU, even for extremely low SpO 2 values, but with no symptoms of fatigue such as dyspnea (Figure 1 ). We compared our 41 patients with a cohort of 24 patients published by Bhatraju et al. 18 . Mean age of admission was comparable between the two studies (median 63 ± 12 vs 64 ± 18 yrs), with patients admitted in Lugano having a slightly lower BMI value (29·3 ± 4·94 vs 33·2 ±7·2, p-val 0.007). No significant difference was found in the hematological status of the two cohorts (WBC, Lymphocytes and Platelets) and comparable level of lactate, maximal CRP level during recovery and liver transaminases (Table 1) pandemic 19, 20 . It requires adequate preparation in terms of hospital structure, triage systems and clinical training 21 in order to be correctly addressed and minimize the burden for the patients and the ICU capacity. As COVID-19 is a multisystemic disease, a multidisciplinary approach is mandatory 4 . To optimize management of tachypnea without dyspnea, also with SpO 2 lower than normal, surveillance tools such as routine and regular Early Warning Score (EWS) measurement (every 4 hours if stable, reduced hourly if SpO 2 < 92%) 11,21 were implemented for intermediate care patients. Patients with COVID-19 interstitial pneumonia present tachypnea correlating with the desaturation degree, without dyspnea or severe neurological . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. . https://doi.org/10.1101/2020.05.03.20089318 doi: medRxiv preprint symptoms or any organ damage. High pulmonary compliance is probably the responsible for the absence of dyspnea 22 ; we waited until the onset of dyspnea or to a value of 85% of SpO 2 before admitting any of the patients to the ICU; in such a way, an overload of the ICU was avoided. None of the 48 patients who after a specialist consultation were not admitted to ICU but only placed under clinical surveillance, died in a follow-up from seven up to 21 days. Most of these patients improved their clinical status without being dyspnoeic and some of them have already been discharged from the hospital. Patients showing a worsening degree of dyspnea transferred to the ICU received "low PEEP ventilatory strategy", on the contrary to what has been reported in the literature 20 . After intubation we found lungs easy to ventilate, with a higher compliance (on average above 50 mL/cmH 2 O) compared to the "classic ARDS" 22 . Even if the "classic" criteria for defining the ARDS were confirmed 23 , there were aspects as the absence of a reduced lung compliance, a "baby-lung" and a consequent tendency to hypercapnia, which induced us to evaluate a more specific treatment, at least in the initial phase. In according to ARDSnet PEEP table 13,14 , we preferred to ventilate patients with PEEP tailored to patients' own BMI, carefully following lungs physiology 23, 24 . This approach would agree with Gattinoni et al 22 and Bendjelid et al 9 , which suggested two different ICU patient populations in COVID-19 pneumonia. The first one presents a high lung compliance and a probable alveolitis, with a shunt effect due to loss of local hypoxic vasoconstriction; this population represents the great majority of our patients. The second one presents a low lung compliance and a picture of baby-lung compatible with "classic ARDS" (only two patients in our set). The "low PEEP ventilatory strategy" we applied allowed us to decrease quickly sedation depth once the inflammation level was reduced. This strategy led to less complications (like ICU paralysis, delayed awakening, agitation, etc..) 18 and an easier and faster extubation without resorting to large-scale tracheotomy. During daily screening of the lower limb ultrasound, COVID-19 ICU patients showed a high prevalence rate of DVTs and PEs, even under preventive anticoagulation. In addition, many patients had a marked increase in Ddimers level, partly linked to the finding of DVT and PE in other sites of the body, partly secondary to PE phenomena also on the pulmonary venous side. In this context, it appears reasonable to protect the patient through a pro-active anticoagulant approach than the normal routine. Furthermore, the fact that patients did not encounter any major bleeding phenomenon, supports the idea that in these patients a more aggressive . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. . anticoagulation may counterbalance a phenomenon of prothrombotic diathesis, even if the complete mechanism is still unclear. We observed an increased survival compared to other groups (Figure 2 and 3) 2,18,24,25 . A possible explanation could be that the relative low-pressure ventilation avoids transforming an initial alveolitis into an ARDSiatrogenic framework, in which the local ongoing inflammation is rather damaged than helped by high PEEP (generating a Ventilation-Induced-Lung Injury -VILI) in a context that is the "not-classic" ARDS. We observed a very few cases of "classic" ARDS and, in particular, the absence of ARDS cases at the time of admission to the ICU. Mortality in ICU is reported to be as high as 42-62% 2,24 , while in our dataset is 24.3%. Median days of MV reported by Bhatraju et al. (8 (2-20) ) is longer (6·5 (3 -10) ) than what experienced in our clinical setting. In all, this suggests that a less traumatic approach to ventilation by low PEEP and avoiding unnecessary MV by delaying ICU admission can be of help in managing COVID-19 patients and in improving survival. Our study was burdened by several limitations. First, it was a monocentric observational retrospective study, with a relatively small series of patients. Second, our comparison with current literature is performed on different patient populations, even if cohorts could be considered similar in terms of disease severity and biochemical investigations. This notwithstanding, early data are very encouraging and needs a validation in bigger prospective studies. In conclusion, the implementation of a multimodal "holistic" approach for COVID-19 patients is highly recommended. We implemented EWS monitoring for intermediate care patients, in order to perform a strict selection of ICU admission and employ MV as little as necessary. MV ventilation was adapted to the real patient needs -i.e. PEEP tailored to patient's BMI -in order to reduce alveolar traumatism. Anticoagulation screening and therapy has been regulated in order to prevent any sign of thrombosis or thromboembolism. This multimodal program allowed us to reduce the number of ICU admissions, the number of ventilation days and mortality, and could be the base for a further specific patients' management in this specific contest. All authors disclose any financial and personal relationships with other people or organizations. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. . Management of the CoViD-19 patients evaluated at the CLM. Patients status admitted to the ICU. At April 12 th , 2020, 10 patients (24.3%) died, 4 (9.7%) are still receiving mechanical ventilation, 2 were transferred to another hospital and 25 (60.9%) were discharged from ICU after a mean of 9 days. All data up to 7 days from extubation have been reported. Overall survival rate during and after ICU admission comparison between the study cohort (class LUG; light blue, continuous line) and the data published by Bhatraju et al 18 (class SEA; light green, dashed line) (p-value = 0·028). . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 8, 2020. ICU mechanical ventilation and laboratory data. Continuous measurements were presented as mean (min-max, ±SD) otherwise as median (IQR) if they are normally distributed. Categorical variables were reported as counts and percentages. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 8, 2020. . 1 8 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 8, 2020. . is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint Coronavirus COVID-19 global cases Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study Critical care utilization for the COVID-19 outbreak in Lombardy, Italy: early experience and forecast during an emergency response Critical care crisis and some recommendations during the COVID-19 epidemic in China Severe SARS-CoV-2 infections: practical considerations and management strategy for intensivists Mechanical Ventilation Redistributes Blood to Poorly Ventilated Areas in Experimental Lung Injury Prevention and treatment of acute lung injury with time-controlled adaptive ventilation: physiologically informed modification of airway pressure release ventilation Covid-19 Does Not Lead to a "Typical" Acute Respiratory Distress Syndrome Treating hypoxiemic patients with SARS-CoV-2 pneumonia. Back to applied physiology COVID-19 pandemic: triage for intensive-care treatment under resource scarcity Management of Adults With Hospital-acquired and Ventilatorassociated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the International ERS/ESICM/ESCMID/ALAT guidelines for the management of hospital-acquired pneumonia and ventilator-associated pneumonia Covid-19 in Critically Ill Patients in the Seattle Region -Case Series Novel coronavirus infection during the 2019-2020 epidemic: preparing intensive care units-the experience in Sichuan Province, China A Comparison of the Quick Sequential (Sepsis-Related) Organ Failure Assessment Score and the National Early Warning Score in Non-ICU Patients With/Without Infection Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention Management of Critically Ill Adults With COVID-19 Critical care crisis and some recommendations during the COVID-19 epidemic in China Less is More" in mechanical ventilation Acute respiratory distress syndrome: the Berlin Definition Clinical Characteristics of Coronavirus Disease 2019 in China Characteristics and Outcomes of 21 Critically Ill Patients With COVID-19 in Washington State JAMA Demographic characteristics and blood tests at admission. Continuous measurements were presented as mean (min-max, ±SD) otherwise as median (IQR) if they are normally distributed. Categorical variables were reported as counts and percentages.