COVID-19-Associated Critical Illness—Report of the First 300 Patients Admitted to Intensive Care Units at a New York City Medical Center

Abstract

Background: The first confirmed case of novel coronavirus (2019-nCoV) infection in the United States was reported from the state of Washington in January, 2020. By March, 2020, New York City had become the epicenter of the outbreak in the United States. Methods: We tracked all patients with confirmed coronavirus-19 (COVID-19) infection admitted to intensive care units (ICU) at Montefiore Medical Center (Bronx, NY). Data were obtained through manual review of electronic medical records. Patients had at least 30 days of follow-up. Results: Our first 300 ICU patients were admitted March 10 through April 11, 2020. The majority (60.7%) of patients were men. Acute respiratory distress syndrome (ARDS) was documented in 91.7% of patients; 91.3% required mechanical ventilation. Prone positioning was employed in 58% of patients and neuromuscular blockade in 47.8% of mechanically-ventilated patients. Neither intervention was associated with decreased mortality. Vasopressors were required in 77.7% of patients. Acute kidney injury (AKI) was present on admission in 40.7% of patients, and developed subsequently in 36.0%; 50.9% of patients with AKI received renal replacement therapy (RRT). Overall 30-day mortality rate was 52.3%, and 55.8% among patients receiving mechanical ventilation. In univariate analysis, higher mortality rate was associated with increasing age, male sex, hypertension, obesity, smoking, number of comorbidities, AKI on presentation, and need for vasopressor support. A representative multivariable model for 30-day mortality is also presented, containing patient age, gender, body mass index, and AKI at admission. As of May 11, 2020, 2 patients (0.7%) remained hospitalized. Conclusions: Mortality in critical illness associated with COVID-19 is high. The majority of patients develop ARDS requiring mechanical ventilation, vasopressor-dependent shock, and AKI. The variation in mortality rates reported to date likely reflects differences in the severity of illness of the evaluated populations.