key: cord-0923612-sijdpep9
authors: Patel, Niti G.; Bhasin, Ajay; Feinglass, Joseph M.; Angarone, Michael P.; Cohen, Elaine R.; Barsuk, Jeffrey H.
title: Mortality, Critical Illness, and Mechanical Ventilation among Hospitalized Patients with COVID-19 on Therapeutic Anticoagulants
date: 2020-12-11
journal: nan
DOI: 10.1016/j.tru.2020.100027
sha: 3deaa5ce43a7aadf2cf964fc2603c95c5da08935
doc_id: 923612
cord_uid: sijdpep9

Background COVID-19 is associated with hypercoagulability and increased incidence of thrombosis. We compared the clinical outcomes of adults hospitalized with COVID-19 who were on therapeutic anticoagulants to those on prophylactic anticoagulation. Materials and Methods We performed an observational study of adult inpatients’ with COVID-19 from March 9 to June 26, 2020. We compared patients who were continued on their outpatient prescribed therapeutic anticoagulation and those who were newly started on therapeutic anticoagulation for COVID-19 (without other indication) to those who were on prophylactic doses. The primary outcome was overall death while secondary outcomes were critical illness (World Health Organization Ordinal Scale for Clinical Improvement score >5), mechanical ventilation, and death among patients who first had critical illness. We adjusted for age, sex, race, body mass index (BMI), Charlson score, glucose on admission, and use of antiplatelet agents. Results Of 1,716 inpatients with COVID-19, 171 patients were continued on their therapeutic anticoagulation and 78 were started on new therapeutic anticoagulation for COVID-19. In patients continued on home therapeutic anticoagulation, there were no differences in overall death, critical illness, mechanical ventilation, or death among patients with critical illness compared to patients on prophylactic anticoagulation. In patients receiving new therapeutic anticoagulation for COVID-19, there was increased death (OR 5.93; 95% CI 3.71-9.47), critical illness (OR 14.51; 95% CI 7.43-28.31), need mechanical ventilation (OR 11.22; 95% CI 6.67-18.86), and death after first having critical illness (OR 5.51; 95% CI 2.80 -10.87). Conclusions Therapeutic anticoagulation for inpatients with COVID-19 was not associated with improved outcomes.

Adults hospitalized with COVID-19 are at increased risk of thromboembolic disease (1) (2) (3) (4) (5) (6) (7) (8) . This has been attributed to COVID-19-associated blood hyperviscosity (9) and vascular endotheliïtis (10) (11) (12) , which suggests that therapeutic strategies directed solely at the coagulation cascade, may be inadequate. A study of an ICU using routine duplex ultrasonography of the lower extremity in critically-ill COVID-19 patients found a high incidence of venous thromboembolism despite therapeutic anticoagulation (13) .

There are no randomized controlled trials and few observational cohort studies addressing the potential benefits and harms of anticoagulating hospitalized patients with COVID-19. Paranjpe et al. reported (14) that therapeutic anticoagulation treatment reduced COVID-19-associated mortality among patients on mechanical ventilation, while Tang et al.

reported (15) that anticoagulation treatment reduced COVID-19-associated mortality among patients with sepsis-induced coagulopathy. However, both of these prior observational studies failed to account for survivorship bias; and the analysis by Tang et al. lacked generalizability as three-quarters of screened patients were excluded from their analysis (16, 17) . Despite the paucity of available evidence, a recent CHEST Guideline and Expert Panel Report recommended that hospitalized COVID-19 patients who are acutely-ill or critically-ill and without a contraindication to anticoagulants be prescribed thromboprophylaxis anticoagulant with lowmolecular-weight heparin (LMWH) or fondaparinux (18) .

In this study, we evaluated the outcomes of adults hospitalized with COVID-19 who either continued therapeutic anticoagulants previously prescribed or who were newly started on anticoagulants during hospitalization. We hypothesized that patients who were prescribed J o u r n a l P r e -p r o o f therapeutic anticoagulants would have better outcomes than those who were on prophylactic anticoagulation while hospitalized.

We performed an observational cohort study of hospitalized adults with COVID-19 from ten hospitals affiliated with Northwestern Medicine between March 9 and June 26, 2020.

Northwestern Medicine is an integrated academic health system in the Chicagoland area. We evaluated clinical outcomes among patients who continued their home therapeutic anticoagulation regimen and those who were newly started on anticoagulants to those who were started on prophylactic anticoagulation during their COVID- were either discharged or died during the study period. We evaluated patients who had been prescribed therapeutic anticoagulation prior to hospitalization and who then continued therapeutic anticoagulation during hospitalization. We also evaluated patients who had not been prescribed therapeutic anticoagulation prior to hospitalization and who then began therapeutic or prophylactic anticoagulation during hospitalization. The anticoagulants specified in the database included those used at Northwestern Medicine: apixaban, argatroban, bivalirudin, dabigatran, dalteparin, edoxaban, enoxaparin, fondaparinux, heparin, rivaroxaban, and warfarin. One author (NGP) manually reviewed the records for the index hospitalization for each patient who received at least one dose of therapeutic anticoagulation to determine the indication for anticoagulation and whether therapeutic anticoagulation had been prescribed prior to admission or was newly started while hospitalized. We categorized patients into the following six groups: continued home therapeutic anticoagulation, started therapeutic anticoagulation for concern of COVID-19 J o u r n a l P r e -p r o o f anticoagulation during hospitalization, or no anticoagulation during hospitalization. We also collected information on age, sex, race, body mass index (BMI), Charlson score, admission blood glucose readings, use of antiplatelet agent (aspirin, clopidogrel, prasugrel, or ticagrelor), ICD-10 codes associated with admission, and outcomes from the WHO Ordinal Scale (Table 2) .

Another author (JHB) performed manual chart review on missing admission glucose levels because the original NUEDW query did not return an admission plasma glucose for a small number of patients. Admission blood glucose was assessed because hyperglycemia is an independent predictor of mortality in patients with COVID-19, regardless of diabetes diagnosis (20) (21) (22) . For patients with multiple admissions, the hospitalization with the worst WHO outcome was analyzed.

We compared patients' prescribed therapeutic anticoagulants (either continued from home or started in the hospital) to prophylactic anticoagulants on four clinical outcomes: overall death (primary outcome), critical illness (as defined as scores ≥5 on the WHO COVID-19 eightpoint ordinal scale), need for mechanical ventilation, and death in critical illness (patients who first had critical illness before dying; WHO score 5-7) (secondary outcomes).

The percentage of overall deaths was calculated as the number of patients who died divided by the total number of discharges and deaths. The percentage of patients with critical illness, mechanical ventilation, and death among patients who first had critical illness was calculated as the number of patients with each outcome divided by the number of critically ill patients who were discharged or died.

We calculated the Charlson comorbidity index for each admission (21, 22) . We used the Chi-square test to determine the significance of differences in baseline demographic and clinical characteristics and clinical outcomes between the six anticoagulation therapy groups. A p-value of <0.01 was used as the threshold for statistical significance in these bivariate analyses to adjust for multiple comparisons. Poisson regression was used to estimate associations between patient attributes and exposures and the likelihood of death, as the incidence rate ratio is a closer approximation to relative risk given the observed 7.3% overall death rate in this cohort (23) .

Multiple logistic regression was used to determine associations between the six anticoagulation therapy groups with the likelihood of critical illness, need for mechanical ventilation and death amongst those with critical illness. Prophylactic anticoagulation was used as the reference group in regression models as this is the standard of care in our hospital. Regression models were adjusted for age, sex, race and ethnicity, body mass index (BMI), Charlson score, glucose on admission, and use of antiplatelet agents. Age, BMI, Charlson score, and blood glucose were categorized for ease of interpretation of odds ratios (24, 25 In bivariate analyses (Table 2) , those patients continued on prior to admission anticoagulation were more likely to be older, have higher Charlson scores, and have more frequent comorbid conditions compared to those beginning prophylactic anticoagulation while hospitalized. Patients newly started on therapeutic anticoagulation compared to those on prophylactic anticoagulation were more likely to have higher Charlson score and more frequent comorbid conditions. Table 2 also shows that patients on therapeutic anticoagulation therapy (both continued outpatient and new inpatient) had worse primary outcomes of overall death and secondary outcomes of critical illness, mechanical ventilation, and death in critically ill compared to those on prophylactic doses. One hundred twenty-five patients in our health system died due to COVID-19 disease (7.3%). Eighteen deaths occurred without critical illness status (WHO score 3-4) and 107 deaths occurred in those who first had critical illness (WHO score 5-7). The mortality rate was 19.3% in critically ill patients.

J o u r n a l P r e -p r o o f 

After adjusting for covariates, there was no difference in death, critical illness, mechanical ventilation, or death in critical illness among patients continued on therapeutic anticoagulation prescribed prior to admission when compared to patients who received prophylactic anticoagulation (Table 3) . Of note, patients who were newly started on therapeutic anticoagulation in the hospital had worse clinical outcomes ( increased Charlson score were associated with worse outcomes (Table 3) . These results were unchanged in the sensitivity analysis.

J o u r n a l P r e -p r o o f (15, 26) . However, neither of these two studies specified whether the dose of anticoagulant was therapeutic or prophylactic. Two other studies limited by low numbers reported lower mortality in patients receiving therapeutic anticoagulants (27, 28) . In contrast, a study of nursing home patients suggested that continuation of oral therapeutic anticoagulation in patients with COVID-19 infection did not have a mortality benefit (29) .

Additionally, another observation study posted on medrxiv.org (that has not yet undergone peer review) reported a higher mortality rate among patients with COVID-19 who received empiric therapeutic anticoagulation with other indications (30) . Based on current evidence, we do not J o u r n a l P r e -p r o o f believe firm recommendations for or against empiric anticoagulation in patients with COVID-19

can be made. In our hospitals, we do not routinely use therapeutic anticoagulation for patients with COVID-19 infection without other indications (31, 32) .

Our study had other significant findings. First, of the patients who were not already on therapeutic anticoagulation, 6.4% of patients with COVID-19 were anticoagulated for a newly diagnosed VTE similar to other studies on the incidence of VTE in COVID-19 (4, 33) . Second, our study confirms previously identified risk factors for worse outcomes in COVID-19 including sex, BMI, blood glucose levels, and comorbid conditions (20, 34, 35) . However, the overall mortality rate (7.3%) of patients in our study was lower than that reported in other studies.

Recent studies out of the New York City area and Louisiana showed inpatient mortality rates between 21-24% (35, 36) . Our mortality rate may be lower because our health system never became overwhelmed, patients were cared for by a multi-disciplinary team that rounded twice daily, and clinicians in our ICUs treated ventilated patients with low volume, high peep strategies as well as prone positioning (37, 38) .

In our cohort, dexamethasone was not widely used during the study period. It seems plausible that routine prophylaxis against or treatment of endotheliïtis with glucocorticosteroids may result in reduced incidence of thrombosis with reductions in overall mortality, as was reported in the RECOVERY study (39, 40) . However, the optimal timing of administration of dexamethasone during COVID-19 illness is not clear as glucocorticosteroids may also suppress innate and adaptive immune response, which may result in higher viral load and greater excursion of blood glucose values, leading to worse outcomes in some patients (41, 42) .

Our study has several limitations. First, our-cohort definition and clinical data relied on the accuracy of ICD-10 coding for disease diagnoses, although we did manually review 372 J o u r n a l P r e -p r o o f (22%) charts which confirmed COVID-19 infection and assessed timing and indication for therapeutic anticoagulation. Second, our study was observational, and despite attempting to control for confounding clinical variables, there may have been confounding by indication. We were unable to determine why therapeutic anticoagulation was started for COVID-19

coagulopathy-whether this was physician preference and/or a marker of more severe illness is unknown. Although patients on therapeutic anticoagulation may have had unmeasured risk factors, we did not find any benefit of therapeutic anticoagulation in patients with COVID-19.

However, in our cohort of patients who were already critically ill, patients started on anticoagulation had 5.5 times the risk of dying. Third, we did not evaluate outcomes by individual anticoagulant drug. It is possible that the clinical effects of DOACs and LMWH differ in patients with COVID-19 (27, 43) . Fourth, we did not control for the number of days patients received new therapeutic anticoagulation because we were unable to determine the timing of starting these medications within the course of COVID-19 infection. Finally, we did not assess the rate of bleeding in our population as it would require a detailed chart review of timing and severity of bleeding. Anticoagulation is a well-known risk factor for bleeding and risk of bleeding may be higher in patients with COVID-19 (33, (44) (45) .

In conclusion, continuation of home therapeutic anticoagulation was associated with neither benefit nor harm in subjects hospitalized with COVID-19. Subjects with COVID-19 who had not previously been anticoagulated and who were empirically prescribed therapeutic anticoagulation during their hospitalization had worse outcomes. Our findings do not support 

Incidence of thrombotic complications in critically ill ICU patients with COVID-19

Extrapulmonary manifestations of COVID-19

Thrombosis in Hospitalized Patients With COVID-19 in a New York City Health System

COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection

Novel insights on the pulmonary vascular consequences of COVID-19

COVID-19-associated coagulopathy: An exploration of mechanisms

Thrombosis risk associated with COVID-19 infection. A scoping review

Venous thromboembolism in non-critically ill patients with COVID-19 infection

COVID-19-associated hyperviscosity: a link between inflammation and thrombophilia?

COVID-19: opening a new paradigm in thromboprophylaxis for critically ill patients?

Pulmonary Vascular Endothelialitis, Thrombosis, and Angiogenesis in Covid-19

High incidence of venous thromboembolic events in anticoagulated severe COVID-19 patients

Association of Treatment Dose Anticoagulation With In-Hospital Survival Among Hospitalized Patients With COVID-19

Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019 patients with coagulopathy

Anticoagulation, immortality, and observations of COVID-19

Thromboprophylaxis for medical inpatients with coronavirus disease 2019

Prevention, Diagnosis, and Treatment of VTE in Patients With Coronavirus Disease

Fasting blood glucose at admission is an independent predictor for 28-day mortality in patients with COVID-19 without previous diagnosis of diabetes: a multi-centre retrospective study

Charlson comorbidity index derived from chart review or administrative data: agreement and prediction of mortality in intensive care patients

New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality

A Modified Poisson Regression Approach to Prospective Studies with Binary Data

Association of All-Cause Mortality With Overweight and Obesity Using Standard Body Mass Index Categories

Diagnosis and Classification of Diabetes Mellitus

The association between treatment with heparin and survival in patients with Covid-19

Protective role of chronic treatment with direct oral anticoagulants in elderly patients affected by interstitial pneumonia in COVID-19 era

A possible benefit from therapeutic anticoagulation in COVID-19: the Dolo hospital experience in

Mortality and the Use of Antithrombotic Therapies Among Nursing Home Residents with COVID-19

Clinical Outcomes With the Use of Prophylactic Versus Therapeutic Anticoagulation in COVID-19. medRxiv

COVID-19 and Thrombotic or Thromboembolic Disease: Implications for Prevention, Antithrombotic Therapy, and Follow-Up: JACC State-of-the-Art Review

COVID-19 and VTE-Anticoagulation -Hematology.org

Incidence of thrombosis and hemorrhage in hospitalized cancer patients with COVID-19

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

Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area

Hospitalization and Mortality among Black Patients and White Patients with Covid-19

Prone Positioning in Awake, Nonintubated Patients With COVID-19 Hypoxemic Respiratory Failure

Surviving Sepsis Campaign: Guidelines on the Management of Critically Ill Adults with Coronavirus Disease 2019 (COVID-19)

Pathophysiology, Transmission, Diagnosis, and Treatment of Coronavirus Disease 2019 (COVID-19): A Review

Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report

Dexamethasone in the management of covid -19

A critical evaluation of glucocorticoids in the management of severe COVID-19

Use of low-molecular-weight heparin in COVID-19 patients

Treatment of Bleeding Complications in Patients on Anticoagulant Therapy

Postdischarge Thrombosis and Hemorrhage in Patients with COVID-19

We would like to acknowledge Dr. Douglas Vaughan for his support of this project and Dr. Brady Stein for editorial feedback on the manuscript. We thank the countless clinicians, environmental and food services personnel, and other staff at Northwestern Medicine for their dedication to patient care during the COVID-19 pandemic.