Risk- and reliability-adjusted estimates for use of hydroxychloroquine (A), tocilizumab (B), prone positioning (C), and neuromuscular blockade (D) across hospitals. Ranking of hospitals differed by treatment modality. Only 35 sites (and 1910 patients) were included in this analysis because the analysis was restricted to sites that submitted data on 15 patients or more. Errors bars indicate 95% CIs.
To convert lymphocytes to ×109/L, multiply by 0.001. BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); ICU, intensive care unit; IMV, invasive mechanical ventilation; Pao2:Fio2, ratio of the Pao2 over the fraction of inspired oxygen; SOFA, Sequential Organ Failure Assessment.
eAppendix 1. STOP-COVID Investigators
eMethods. Supplemental Methods
eTable 1. List of Participating Sites
eTable 2. Definitions of Baseline Characteristics, Comorbidities, Treatments, and Outcomes
eTable 3. Complete List of Patient Characteristics at Baseline
eTable 4. Frequency of Symptoms, Medications, Comorbidities
eTable 5. Frequency of Other Complications
eTable 6. Inter-hospital Variation in Treatments and Outcomes
eFigure 1. Flowchart for Site Selection
eFigure 2. Number of Cohort Patients by City Among Contributing Sites
eFigure 3. Frequency Distribution of Patient and Hospital Characteristics and 28-Day Mortality
eFigure 4. Frequency Distribution of Laboratory and Physiologic Parameters on Admission to the ICU and 28-Day Mortality
eFigure 5. Acute Organ Injury
eFigure 6. Temporal Changes in Laboratory Values and Physiologic Parameters
eFigure 7. 28-Day Mortality by Geographic Region
eFigure 8. Inter-hospital Variation in Mortality
eFigure 9. Multivariable Risk Model for Death at 28 Days in Mechanically Ventilated Patients
eFigure 10. Multivariable Risk Model for Death using Time-to-Event Analyses
eAppendix 2. Case Report Form
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Gupta S, Hayek SS, Wang W, et al. Factors Associated With Death in Critically Ill Patients With Coronavirus Disease 2019 in the US. JAMA Intern Med. 2020;180(11):1436–1447. doi:10.1001/jamainternmed.2020.3596
What are the characteristics, outcomes, and factors associated with death among critically ill patients with coronavirus disease 2019 (COVID-19) in the US?
In a cohort of 2215 adults with COVID-19 who were admitted to intensive care units at 65 sites, 784 (35.4%) died within 28 days, with wide variation among hospitals. Factors associated with death included older age, male sex, morbid obesity, coronary artery disease, cancer, acute organ dysfunction, and admission to a hospital with fewer intensive care unit beds.
This study identified demographic, clinical, and hospital-level factors associated with death in critically ill patients with COVID-19 that may be used to facilitate the identification of medications and supportive therapies that can improve outcomes.
The US is currently an epicenter of the coronavirus disease 2019 (COVID-19) pandemic, yet few national data are available on patient characteristics, treatment, and outcomes of critical illness from COVID-19.
To assess factors associated with death and to examine interhospital variation in treatment and outcomes for patients with COVID-19.
Design, Setting, and Participants
This multicenter cohort study assessed 2215 adults with laboratory-confirmed COVID-19 who were admitted to intensive care units (ICUs) at 65 hospitals across the US from March 4 to April 4, 2020.
Patient-level data, including demographics, comorbidities, and organ dysfunction, and hospital characteristics, including number of ICU beds.
Main Outcomes and Measures
The primary outcome was 28-day in-hospital mortality. Multilevel logistic regression was used to evaluate factors associated with death and to examine interhospital variation in treatment and outcomes.
A total of 2215 patients (mean [SD] age, 60.5 [14.5] years; 1436 [64.8%] male; 1738 [78.5%] with at least 1 chronic comorbidity) were included in the study. At 28 days after ICU admission, 784 patients (35.4%) had died, 824 (37.2%) were discharged, and 607 (27.4%) remained hospitalized. At the end of study follow-up (median, 16 days; interquartile range, 8-28 days), 875 patients (39.5%) had died, 1203 (54.3%) were discharged, and 137 (6.2%) remained hospitalized. Factors independently associated with death included older age (≥80 vs <40 years of age: odds ratio [OR], 11.15; 95% CI, 6.19-20.06), male sex (OR, 1.50; 95% CI, 1.19-1.90), higher body mass index (≥40 vs <25: OR, 1.51; 95% CI, 1.01-2.25), coronary artery disease (OR, 1.47; 95% CI, 1.07-2.02), active cancer (OR, 2.15; 95% CI, 1.35-3.43), and the presence of hypoxemia (Pao2:Fio2<100 vs ≥300 mm Hg: OR, 2.94; 95% CI, 2.11-4.08), liver dysfunction (liver Sequential Organ Failure Assessment score of 2-4 vs 0: OR, 2.61; 95% CI, 1.30–5.25), and kidney dysfunction (renal Sequential Organ Failure Assessment score of 4 vs 0: OR, 2.43; 95% CI, 1.46–4.05) at ICU admission. Patients admitted to hospitals with fewer ICU beds had a higher risk of death (<50 vs ≥100 ICU beds: OR, 3.28; 95% CI, 2.16-4.99). Hospitals varied considerably in the risk-adjusted proportion of patients who died (range, 6.6%-80.8%) and in the percentage of patients who received hydroxychloroquine, tocilizumab, and other treatments and supportive therapies.
Conclusions and Relevance
This study identified demographic, clinical, and hospital-level risk factors that may be associated with death in critically ill patients with COVID-19 and can facilitate the identification of medications and supportive therapies to improve outcomes.
Since the outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection began in December 2019 in Wuhan, China, more than 6 million people have developed coronavirus disease 2019 (COVID-19), and more than 350 000 have died.1 Critical illness from COVID-19 in China, Italy, and other countries has strained intensive care unit (ICU) resources and produced a wide spectrum of short-term mortality rates, ranging from 16% to 62%.2-4
As of June 19, 2020, approximately 2.2 million people in the US have been infected with SARS-CoV-2, and more than 100 000 have died. Although more people have died in the US than in any other country,1 national data are lacking on the epidemiologic factors, treatment, and outcomes of critical illness from COVID-19. One study5 of 24 patients in the Seattle, Washington, region reported frequent receipt of invasive mechanical ventilatory support and vasopressors and an in-hospital mortality of 50%. Local outbreaks of COVID-19 in New York City have been described in single-center and regional reports.6,7 These studies6,7 included primarily noncritically ill patients and had limited follow-up duration.
Granular data on patient characteristics, treatment, and outcomes of critical illness from COVID-19 are needed to inform decision-making about resource allocation, critical care capacity, and treatment of patients. Furthermore, nationally representative data across multiple hospitals are needed to assess interhospital variation in treatment and outcomes. To address this knowledge gap, we conducted the Study of the Treatment and Outcomes in Critically Ill Patients With COVID-19 (STOP-COVID), a multicenter cohort study that examined the demographics, comorbidities, organ dysfunction, treatment, and outcomes of patients with COVID-19 admitted to ICUs across the US. The purposes of this study were to assess factors associated with death and to examine interhospital variation in treatment and outcomes in patients with COVID-19.
In this multicenter cohort study, we enrolled adults with COVID-19 who were admitted to participating ICUs at 65 hospitals. The study was approved by the institutional review boards at each participating site with a waiver of informed consent. All data except dates were deidentified.
We included consecutive adult patients (≥18 years of age) with laboratory-confirmed COVID-19 (detected by nasopharyngeal or oropharyngeal swab) admitted to a participating ICU for illness related to COVID-19 between March 4 and April 4, 2020. Patients were considered to have been admitted to an ICU if they were admitted to a regular ICU or if they were in a non-ICU room that was functioning as an ICU room for surge capacity (defined further in the eMethods and eAppendix in Supplement 1). We followed up patients until hospital discharge, death, or June 4, 2020, whichever came first. A complete list of participating sites is provided in eTable 1 and eFigures 1 and 2 in Supplement 1. In this cohort, 98 patients were described in prior studies, including 58 reported in a single-center case series from New York City that included both critically ill and noncritically ill patients,6 28 reported in a single-center study focused on acute kidney injury,8 and 12 reported in a case series of critically ill patients from the Seattle region.5
The primary outcome was death within 28 days of ICU admission. Patients who were discharged alive from the hospital before 28 days were considered to be alive at 28 days (we tested the validity of this assumption in a subset of patients, described in the eMethods in Supplement 1). Secondary outcomes included development of respiratory failure, acute respiratory distress syndrome, congestive heart failure, myocarditis, pericarditis, arrhythmia, shock, acute cardiac injury, acute kidney injury, acute liver injury, coagulopathy, secondary infection, and thromboembolic events (definitions provided in eTable 2 in Supplement 1). We also examined receipt of antivirals, antibiotics, anticoagulants, immunomodulating medications, mechanical ventilatory support, adjunctive and rescue therapies for hypoxemia, extracorporeal membrane oxygenation, mechanical cardiac support, vasopressors, and kidney replacement therapy.
Study personnel at each site collected data by manual review of electronic medical records and used a standardized case report form to enter data into a secure online database (eAppendix 2 in Supplement 1). Patient-level data included baseline information on demographics, coexisting conditions, symptoms, medications before hospital admission, and vital signs; daily data for the 14 days after ICU admission on physiologic and laboratory values, medications, nonmedication treatments, and organ support; and outcome data on ICU and hospital length of stay and death. We also collected hospital-level data, including the city and state, main hospital vs satellite or affiliate hospital for the center, the number of ICU beds (not including surge capacity), and type of ICU to which the patients were admitted (medical, COVID specific, or other). All data were validated through a series of automated and manual verifications (eMethods in Supplement 1).
We aimed to generate a representative sample of critically ill adults by enrolling at least 2000 patients from at least 50 geographically diverse hospitals. To describe baseline characteristics, treatment, and outcomes, we express continuous variables as median (interquartile range [IQR]) and categorical variables as number (percentage).
To assess interhospital variation in treatments and outcomes, we used multilevel conditional logistic regression modeling with patients nested in hospitals to characterize hospital-level variation in treatment and to estimate hospital-specific rates of death at 28 days. This approach addresses the poor reliability of estimates stemming from hospitals that submitted few cases. To further improve the reliability of the estimates, we excluded hospitals that submitted data on fewer than 15 patients.
To account for differences in patient-level characteristics and illness severity among hospitals, we prespecified the following covariates for inclusion in the models: age, sex, race, hypertension, diabetes, body mass index (calculated as weight in kilograms divided by height in meters squared), coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease, current smoking status, active cancer, duration of symptoms before ICU admission, and covariates assessed at ICU admission (lymphocyte count, ratio of the Pao2 to the fraction of inspired oxygen [Fio2], shock, and the kidney, liver, and coagulation components of the Sequential Organ Failure Assessment score).9 We included the above characteristics and patient-level deviations from hospital means in the models. Finally, in a sensitivity analysis, we further adjusted for the number of ICU beds at each hospital before the COVID-19 pandemic (<50, 50-99, and ≥100 ICU beds).
To identify independent associations between patient and hospital characteristics and the primary outcome of 28-day mortality, we again used multilevel logistic regression modeling, with a random effect for hospital and fixed effects for each variable of interest. We included the same covariates as above. We also performed 2 sensitivity analyses. First, we repeated the above analysis but limited the patients to those who received invasive mechanical ventilatory support on ICU day 1. Second, we used a Cox regression model and all available follow-up data to assess the association between the above covariates and survival.
Data regarding body mass index were missing for 109 patients (4.9%). Data regarding the ratio of Pao2:Fio2 at ICU admission were missing for 273 of 1494 patients (18.3%). Additional details on missing data are provided in the eMethods and eTable 3 in Supplement 1. Analyses were performed using SAS software, version 9.4 (SAS Institute Inc) and Stata, version 16.1 (StataCorp LLC).
From the 65 sites, 2215 of 2833 patients initially considered met eligibility criteria and were included in the analysis. The mean (SD) age was 60.5 (14.5) years, and 1436 (64.8%) were men. The median duration of symptoms before ICU admission was 7 days (IQR, 4-10 days). The most common symptoms before ICU admission were cough (1708 [77.1%]), dyspnea (1658 [74.9%]), and fever (1566 [70.7%]). A total of 1738 patients (78.5%) had at least 1 coexisting condition, including hypertension (1322 [59.7%]), diabetes (861 [38.9%]), and chronic lung disease (531 [24.0%]). On the day of ICU admission, 1494 patients (67.4%) received invasive mechanical ventilatory support, and 958 (48.3%) received vasopressors. The median Pao2:Fio2 ratio was 124 mm Hg (IQR, 86-188 mm Hg). Additional characteristics are provided in Table 1 and eTables 3 and 4 and eFigures 3 and 4 in Supplement 1.
In the 14 days after ICU admission, 1859 patients (83.9%) received invasive mechanical ventilatory support, 1635 patients (73.8%) developed acute respiratory distress syndrome, and 921 of the 2151 patients without end-stage kidney disease (42.8%) developed acute kidney injury. Other acute organ injuries were less frequent in this study, with only 230 patients (10.4%) experiencing a clinically detected thromboembolic event. The incidence of other acute organ injuries is shown in eFigure 5 in Supplement 1.
The median time from symptom onset to each acute organ injury is shown in eFigure 5 in Supplement 1. Respiratory failure, acute cardiac injury, and congestive heart failure occurred earlier in the illness, whereas secondary infection, acute liver injury, and thromboembolic events occurred later (eTable 5 in Supplement 1). Longitudinal assessment of laboratory values and physiologic parameters is shown in eFigure 6 in Supplement 1.
The most commonly administered medications for COVID-19–related illness were hydroxychloroquine (1761 [79.5%]), azithromycin (1320 [59.6%]), and therapeutic anticoagulants (920 [41.5%]). Interventions for hypoxemia included neuromuscular blockade (909 [41.0%]), prone positioning (852 [38.5%]), inhaled epoprostenol (118 [5.3%]), and inhaled nitric oxide (94 [4.2%]). Additional medications and supportive therapies are listed in Table 2.
Risk- and reliability-adjusted use of medications and supportive therapies varied widely among hospitals (Figure 1 and eTable 6 in Supplement 1). For example, the risk- and reliability-adjusted proportion of patients who received hydroxychloroquine was 82.2% overall but ranged from 16.8% at the lowest use hospital to 98.1% at the highest. Similarly, the risk- and reliability-adjusted proportion of patients who received prone positioning was 35.1% overall but ranged from 4.6% at the lowest use hospital to 79.9% at the highest. Considerable interhospital variation in the use of therapies persisted in analyses further adjusted for the number of ICU beds (eTable 6 in Supplement 1).
Overall, 784 patients (35.4%) died within 28 days of ICU admission, 824 (37.2%) were discharged alive from the hospital within 28 days, and 607 (27.4%) remained hospitalized at 28 days. The unadjusted incidence of death within 28 days of ICU admission is displayed according to baseline patient and hospital characteristics in eFigures 3 and 4 in Supplement 1 and according to geographic region in eFigure 7 in Supplement 1. The most common causes of death were respiratory failure (727 [92.7%]), septic shock (311 [39.7%]), and kidney failure (295 [37.6%]), with many patients having more than 1 cause.
Risk- and reliability-adjusted rate of death within 28 days varied widely across hospitals, from 6.6% to 80.8% (eFigure 8 and eTable 6 in Supplement 1). After further adjustment for the number of ICU beds before the pandemic, this variation decreased, such that the risk- and reliability-adjusted rate of death ranged from 11.9% to 63.3% (eTable 6 in Supplement 1).
In a multivariable model that examined the association between prespecified patient and hospital characteristics and 28-day mortality, older age was independently associated with higher risk of death (≥80 vs <40 years of age: odds ratio [OR], 11.15; 95% CI, 6.19-20.06) (ORs for other age categories are shown in Figure 2). Additional characteristics associated with death were male sex (OR, 1.50; 95% CI, 1.19-1.90), higher body mass index (≥40 vs <25: OR, 1.51; 95% CI, 1.01-2.25), coronary artery disease (OR, 1.47; 95% CI, 1.07-2.02), active cancer (OR, 2.15; 95% CI, 1.35-3.43), and the presence of hypoxemia (Pao2:Fio2<100 vs ≥300 mm Hg: OR, 2.94; 95% CI, 2.11-4.08), liver dysfunction (liver Sequential Organ Failure Assessment score of 2-4 vs 0: OR, 2.61; 95% CI, 1.30–5.25), and kidney dysfunction (renal Sequential Organ Failure Assessment score of 4 vs 0: OR, 2.43; 95% CI, 1.46–4.05) at ICU admission (Figure 2). Race (race other than White vs White race: OR, 1.11; 95% CI, 0.88-1.40), hypertension (OR, 1.06; 95% CI, 0.83-1.36), diabetes (OR, 1.14; 0.91-1.43), and lymphocyte count (OR, 1.11; 95% CI, 0.88-1.41) were not associated with death. Patients admitted to hospitals with fewer ICU beds had a higher risk of death (<50 vs ≥100 ICU beds; OR, 3.28; 95% CI, 2.16-4.99) (Figure 2). Interpretations were largely unchanged when restricted to patients who received invasive mechanical ventilatory support on ICU day 1 (eFigure 9 in Supplement 1), although a body mass index greater than 40 was no longer associated with a higher risk of death.
Among patients discharged alive from the hospital within 28 days of ICU admission, the median ICU length of stay was 9 days (IQR, 5-14 days) and the median hospital length of stay was 16 days (IQR, 11-22 days). Extracorporeal organ support included acute kidney replacement therapy (432 [20.1%]), extracorporeal membrane oxygenation (61 [2.8%]), and mechanical cardiac support (3 [0.1%]). Additional outcomes are provided in Table 3.
At the end of study follow-up (median, 16 days; interquartile range, 8-28 days), a total of 875 patients (39.5%) had died, 1203 (54.3%) were discharged alive from the hospital, and 137 (6.2%) remained hospitalized. Among patients discharged alive from the hospital, the median ICU length of stay was 12 days (IQR, 6-21 days), and the median hospital length of stay was 21 days (IQR, 13-33 days). Using a multivariable Cox regression model, we identified factors associated with death that were similar to those in the 28-day mortality model (eFigure 10 in Supplement 1).
This multicenter cohort study of 2215 critically ill adults with COVID-19 admitted to ICUs at 65 hospitals across the US found that 784 patients (35.4%) died in the 28 days after ICU admission. Older age, male sex, higher body mass index, coronary artery disease, and active cancer were independently associated with a higher risk of death, as was the presence of hypoxemia and liver and kidney dysfunction at ICU admission. Patients admitted to hospitals with fewer ICU beds also had a higher risk of death. Hospitals varied widely in the proportion of patients who received medications and supportive therapy for COVID-19 and in the proportion of patients who died.
Prior data on critical illness from COVID-19 derive from cohorts in China and Italy and small case series and regional reports from cohorts in the US.3-6 Compared with a large cohort of critically ill patients with COVID-19 in Lombardy, Italy, the median age of patients in the cohort in the present study and the proportion who received invasive mechanical ventilatory support were similar.4 The mortality in the cohort in the present study was higher than that of critically ill patients with COVID-19 in Italy (26%),4 although 58% of the patients in that cohort were still in the ICU at the end of follow-up, but lower than that reported in single-center studies from Wuhan, China (62%)3 and the Seattle region of the US (50%).5 These comparisons are limited by different ICU admitting practices and duration of follow-up among studies.
The most common acute organ injuries observed in the cohort in this study were respiratory failure, acute respiratory distress syndrome, and acute kidney injury. Other acute organ injuries were less frequent in this study, with only 230 patients (10.4%) experiencing a clinically detected thromboembolic event. This incidence of thromboembolic events is considerably lower than the 15% to 42% incidence reported in critically ill patients with COVID-19 in Europe10-14 and is more consistent with the incidence reported in critically ill patients without COVID-19.15 Understanding the reason for these differences will be important as hypercoagulability in COVID-19 is pursued as a potential therapeutic target.16
This study identified considerable interhospital variation in the administration of medications and supportive therapies intended to treat COVID-19 and associated organ injury. Sources of this variation may include the limited high-quality evidence on which to base clinical practice, variation in hospital resources to implement personnel-intensive interventions (eg, prone positioning), variation in the availability of certain medications (eg, remdesivir), or unmeasured variation in patient and practitioner characteristics across centers. These data support clinical equipoise for ongoing randomized clinical trials of therapies for COVID-19.
In this study, several patient characteristics were associated with a higher risk of death. Similar to previous reports,17,18 older age was associated with a higher risk of death, although at least 15% of patients died in every age group, including those younger than 40 years. Two-thirds of the patients were men, and male sex was independently associated with a higher risk of death, supporting a prior report19 of the association between male sex and adverse outcomes in patients with COVID-19. In addition, we found that higher body mass index was independently associated with a higher risk of death, extending the findings from prior reports20-22 on the association between obesity and severe illness from COVID-19. We also identified several novel patient-level factors associated with death, including coronary artery disease and active cancer. Finally, we found that patients who were admitted to hospitals with fewer ICU beds had a higher risk of death.
Nearly 1 in 3 patients in the present cohort was Black compared with approximately 13.4% of the US population. Race, however, was not associated with death in multivariable models. These results are similar to those recently reported by a single-center study23 in Louisiana, which found that Black patients were more likely to be hospitalized but had similar in-hospital mortality compared with White patients. The reasons for potential racial differences in the frequency of ICU admission with COVID-19 are likely multifactorial and may reflect differences in comorbidities, socioeconomic status, and other factors.
This study has several strengths. First, we collected comprehensive data from a large number of consecutive critically ill patients with laboratory-confirmed COVID-19 for 28 days, thereby minimizing selection or surveillance bias at each center. Second, we included patients from 65 geographically diverse sites from across the US, thereby maximizing generalizability. By including a large number of hospitals, we also identified considerable variation in risk-adjusted practice patterns and outcomes across sites. Third, we obtained all data by detailed medical record review rather than reliance on administrative or billing codes, which have well-described limitations.24 Fourth, whereas prior studies4-7 followed up patients for shorter periods, we followed up patients until the first occurrence of hospital discharge, death, or 28 days in our primary analyses and for up to 3 months in secondary analyses. By including additional follow-up, we were able to ascertain a definitive in-hospital mortality outcome (death or discharged) in 93.8% of the patients in our cohort. With the additional follow-up, the in-hospital mortality rate was 39.5%.
We acknowledge several limitations. Although health care center was included as a covariate in multivariable models, there may be unmeasured differences in patient populations among hospitals, explaining some of the observed variation in treatments and outcomes. Although some acute organ injuries (eg, acute kidney, liver, and cardiac injury) were assessed by objective laboratory-based definitions, others (eg, acute respiratory distress syndrome) relied on the clinical impression of the treating practitioner and may have differed from the true incidence. Although we assessed mortality and length of stay for 28 days, we assessed laboratory and physiologic parameters, acute organ injury, and organ support for the first 14 days only after ICU admission.
Our estimates of interhospital variation in risk- and reliability-adjusted rates of death may be affected by residual confounding because of differences in baseline risk and patient and physician characteristics across hospitals that were not accounted for by our measured covariates. For example, other than data on homelessness (0.6% prevalence in our cohort), we did not collect data on the socioeconomic status of the patients. Socioeconomic status is increasingly recognized as an important factor associated with health outcomes in patients with COVID-1925 and could have influenced our findings with respect to variation in mortality across hospitals. We also did not collect detailed data on ventilator management strategies, hospital or ICU patient volume, or physician and nurse availability. Our models do not account for varying degrees of strain on the available resources across hospitals, such as the extent to which it may have increased the number of ICU beds beyond the baseline number before the pandemic. We did not collect data on do-not-resuscitate or do-not-intubate orders or the availability of palliative care for patients. These factors may have contributed to differing rates of intubation and ICU admission, and thus mortality, across centers. Accordingly, our findings should be interpreted cautiously. Further studies should build on our findings and seek to better understand the reasons for the considerable interhospital variation in outcomes that we observed.
In this multicenter cohort study of critically ill adults with COVID-19 in the US, more than 1 in 3 died within 28 days after ICU admission. We identified several patient- and hospital-level factors that were associated with death and found that treatment and outcomes varied considerably among hospitals. Future research should examine the patients with COVID-19 at greatest risk of adverse outcomes and seek to identify medications or supportive therapies that improve their outcomes.
Accepted for Publication: June 19, 2020.
Published Online: July 15, 2020. doi:10.1001/jamainternmed.2020.3596
Correction: This article was corrected on August 6, 2020, to correct an error in the Key Points and to list the STOP-COVID Investigators in the Article Information. The article was also corrected on September 10, 2020, to remove the incorrect expansion of the BREATHE trial in the conflict of interest disclosures. The article was also corrected on April 26, 2021, to correct included measures of SOFA scores and to add an additional Supplement listing the names of the STOP-COVID Investigators, and on June 14, 2021, to restore eAppendix 2 to Supplement 1.
Corresponding Author: David E. Leaf, MD, MMSc, Division of Renal Medicine, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115 (firstname.lastname@example.org).
Author Contributions: Drs Gupta and Leaf had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Gupta, Hayek, Melamed, Radbel, Green, Shaefi, Parikh, Arunthamakun, Kibbelaar, Gershengorn, Leaf.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Gupta, Hayek, Wang, Mathews, Radbel, Reiser, Green, Vijayan, Abu Omar, Admon, Semler, Leaf.
Critical revision of the manuscript for important intellectual content: Gupta, Hayek, Wang, Chan, Mathews, Melamed, Brenner, Leonberg-Yoo, Schenck, Radbel, Bansal, Srivastava, Zhou, Sutherland, Green, Shehata, Goyal, Velez, Shaefi, Parikh, Arunthamakun, Athavale, Friedman, Short, Kibbelaar, Admon, Donnelly, Gershengorn, Hernán, Semler, Leaf.
Statistical analysis: Gupta, Wang, Mathews, Admon, Donnelly, Hernán, Semler, Leaf.
Administrative, technical, or material support: Gupta, Hayek, Chan, Schenck, Radbel, Reiser, Bansal, Zhou, Goyal, Velez, Shaefi, Parikh, Arunthamakun, Athavale, Short, Kibbelaar, Abu Omar, Leaf.
Supervision: Hayek, Mathews, Leonberg-Yoo, Zhou, Sutherland, Shaefi, Arunthamakun, Athavale, Hernán, Semler, Leaf.
Conflict of Interest Disclosures: Dr Gupta reported receiving grants from the National Institutes of Health (NIH) and is a scientific coordinator for GlaxoSmithKline’s ASCEND (Anemia Studies in Chronic Kidney Disease: Erythropoiesis via a Novel Prolyl Hydroxylase Inhibitor Daprodustat) trial. Dr Chan reported receiving grants from the Renal Research Institute outside the submitted work. Dr Mathews reported receiving grants from the NIH/National Heart, Lung, and Blood Institute (NHLBI) during the conduct of the study and serves on the steering committee for the BREATHE trial, funded by Roivant/Kinevant Sciences. Dr Melamed reported receiving honoraria from the American Board of Internal Medicine and Icon Medical Consulting. Dr Reiser reported receiving personal fees from Biomarin, TRISAQ, Thermo BCT, Astellas, Massachusetts General Hospital, Genentech, UptoDate, Merck, Inceptionsci, GLG, and Clearview and grants from the NIH and Nephcure outside the submitted work. Dr Srivastava reported receiving personal fees from Horizon Pharma PLC, AstraZeneca, and CVS Caremark outside the submitted work. Dr Vijayan reported receiving personal fees from NxStage, Boeringer Ingelheim, and Sanofi outside the submitted work. Dr Velez reported receiving personal fees from Mallinckrodt Pharmaceuticals, Retrophin, and Otsuka Pharmaceuticals outside the submitted work. Dr Shaefi reported receiving grants from the NIH/National Institute on Aging and NIH/National Institute of General Medical Sciences outside the submitted work. Dr Admon reported receiving grants from the NIH/NHLBI during the conduct of the study. Dr Donnelly reported receiving grants from the NIH/NHLBI during the conduct of the study and personal fees from the American College of Emergency Physicians/Annals of Emergency Medicine outside the submitted work. Dr Hernán reported receiving grants from the NIH during the conduct of the study. Dr Semler reported receiving grants from the NIH/NHLBI during the conduct of the study. No other disclosures were reported.
Group Information: The STOP-COVID Investigators are listed in eAppendix 1 in Supplement 1 and in Supplement 2. They include the following: Carl P. Walther, Samaya J. Anumudu, Kathleen F. Kopecky, Gregory P. Milligan, Peter A. McCullough, Thuy-Duyen Nguyen, Megan L. Krajewski, Sidharth Shankar, Ameeka Pannu, Juan D. Valencia, Sushrut S. Waikar, Peter Hart, Oyintayo Ajiboye, Matthew Itteera, Jean-Sebastien Rachoin, Christa A. Schorr, Lisa Shea, Daniel L. Edmonston, Christopher L. Mosher, Aaron Karp, Zaza Cohen, Valerie Allusson, Gabriela Bambrick-Santoyo, Noor ul aain Bhatti, Bijal Mehta, Aquino Williams, Patricia Walters, Ronaldo C. Go, Keith M. Rose, Amy M. Zhou, Ethan C. Kim, Rebecca Lisk, Steven G. Coca, Deena R. Altman, Aparna Saha, Howard Soh, Huei Hsun Wen, Sonali Bose, Emily A. Leven, Jing G. Wang, Gohar Mosoyan, Girish N. Nadkarni, John Guirguis, Rajat Kapoor, Christopher Meshberger, Brian T. Garibaldi, Celia P. Corona-Villalobos, Yumeng Wen, Steven Menez, Rubab F. Malik, Carmen Elena Cervantes, Samir C. Gautam, H. Bryant Nguyen, Afshin Ahoubim, Leslie F. Thomas, Dheeraj Reddy Sirganagari, Pramod K. Guru, Paul A. Bergl, Jesus Rodriguez, Jatan A. Shah, Mrigank S. Gupta, Princy N. Kumar, Deepa G. Lazarous, Seble G. Kassaye, Tanya S. Johns, Ryan Mocerino, Kalyan Prudhvi, Denzel Zhu, Rebecca V. Levy, Yorg Azzi, Molly Fisher, Milagros Yunes, Kaltrina Sedaliu, Ladan Golestaneh, Maureen Brogan, Ritesh Raichoudhury, Soo Jung Cho, Maria Plataki, Sergio L. Alvarez-Mulett, Luis G. Gomez-Escobar, Di Pan, Stefi Lee, Jamuna Krishnan, William Whalen, David Charytan, Ashley Macina, Daniel W. Ross, Alexander S. Leidner, Carlos Martinez, Jacqueline M. Kruser, Richard G. Wunderink, Alexander J. Hodakowski, Eboni G. Price-Haywood, Luis A. Matute-Trochez, Anna E. Hasty, Muner MB. Mohamed, Rupali S. Avasare, David Zonies, Rebecca M. Baron, Meghan E. Sise, Erik T. Newman, Kapil K. Pokharel, Shreyak Sharma, Harkarandeep Singh, Simon Correa, Tanveer Shaukat, Omer Kamal, Heather Yang, Jeffery O. Boateng, Meghan Lee, Ian A. Strohbehn, Jiahua Li, Saif A. Muhsin, Ernest I. Mandel, Ariel L. Mueller, Nicholas S. Cairl, Chris Rowan, Farah Madhai-Lovely, Vasil Peev, John J. Byun, Andrew Vissing, Esha M. Kapania, Zoe Post, Nilam P. Patel, Joy-Marie Hermes, Amee Patrawalla, Diana G. Finkel, Barbara A. Danek, Sowminya Arikapudi, Jeffrey M. Paer, Sonika Puri, Jag Sunderram, Matthew T. Scharf, Ayesha Ahmed, Ilya Berim, Sabiha Hussain, Shuchi Anand, Joseph E. Levitt, Pablo Garcia, Suzanne M. Boyle, Rui Song, Jingjing Zhang, Moh’d A. Sharshir, Vadym V. Rusnak, Amber S. Podoll, Michel Chonchol, Sunita Sharma, Ellen L. Burnham, Arash Rashidi, Rana Hejal, Erik T. Judd, Laura Latta, Ashita Tolwani, Timothy E. Albertson, Jason Y. Adams, Steven Y. Chang, Rebecca M. Beutler, Carl E. Schulze, Etienne Macedo, Harin Rhee, Kathleen D. Liu, Vasantha K. Jotwani, Jay L. Koyner, Chintan V. Shah, Vishal Jaikaransingh, Stephanie M. Toth-Manikowski, Min J. Joo, James P. Lash, Javier A. Neyra, Nourhan Chaaban, Alfredo Iardino, Elizabeth H. Au, Jill H. Sharma, Marie Anne Sosa, Sabrina Taldone, Gabriel Contreras, David De La Zerda, Pennelope Blakely, Hanna Berlin, Tariq U. Azam, Husam Shadid, Michael Pan, Patrick O’ Hayer, Chelsea Meloche, Rafey Feroze, Kishan J. Padalia, Abbas Bitar, Jennifer E. Flythe, Matthew J. Tugman, Brent R. Brown, Ryan C. Spiardi, Todd A. Miano, Meaghan S. Roche, Charles R. Vasquez, Amar D. Bansal, Natalie C. Ernecoff, Csaba P. Kovesdy, Miklos Z. Molnar, Ambreen Azhar, Susan S. Hedayati, Mridula V. Nadamuni, Sadaf S. Khan, Duwayne L. Willett, Amanda D. Renaghan, Pavan K. Bhatraju, Bilal A. Malik, Christina Mariyam Joy, Tingting Li, Seth Goldberg, Patricia F. Kao, Greg L. Schumaker, Anthony J. Faugno, Caroline M. Hsu, Asma Tariq, Leah Meyer, Daniel E. Weiner, Marta Christov, Francis P. Wilson, Tanima Arora, Ugochukwu Ugwuowo, Erik T. Newman.
Additional Contributions: Dino Mazzarelli, JD, Partners Healthcare Research Management, Boston, Massachusetts, and Patricia Reaser, Division of Renal Medicine at Brigham and Women’s Hospital, Boston, Massachusetts, provided assistance coordinating the data use agreements with each institution. No compensation was provided to these individuals. We thank the clinical and research staff from each of the participating sites.