Abstract
Objective: To characterize patients with COVID-19 in a large New York City (NYC) medical center and describe their clinical course across the emergency department (ED), inpatient wards, and intensive care units (ICUs). Design: Retrospective manual medical record review. Setting: NewYork-Presbyterian/Columbia University Irving Medical Center (NYP/CUIMC). Participants: The first 1000 consecutive patients with laboratory-confirmed COVID-19. Methods: Manual abstraction of 274 data fields from electronic health records (EHR). Results: Among the 1000 patients, 150 were ED patients, 614 were admitted without ICU care, and 236 required ICU care. Common symptoms included cough (73.2%), fever (72.8%), and dyspnea (63.1%). ICU patients were older, predominantly male (66.9%), and had long stays (median 23 days). Notably, 78.0% of ICU patients developed AKI. Time to intubation showed a bimodal distribution with peaks at 3-4 and 9 days. Conclusions: Hospitalized COVID-19 patients faced significant morbidity and mortality, with high rates of AKI and a distinct bimodal distribution in the timing of respiratory failure.
Introduction
In early 2020, New York City emerged as the global epicenter of the COVID-19 pandemic. While early data from Wuhan and Lombardy provided initial insights, there was a critical need for granular, U.S.-based data regarding patient characteristics, clinical progression, and resource utilization. Clinicians needed to understand who was most at risk and when patients were most likely to deteriorate. This study addresses these questions by providing a detailed analysis of the first 1,000 consecutive adult patients treated at NewYork-Presbyterian/Columbia University Irving Medical Center (NYP/CUIMC).
Methods
The researchers conducted a retrospective manual medical record review of the first 1,000 consecutive adult patients with laboratory-confirmed SARS-CoV-2 who presented to the ED or were hospitalized between March 1 and April 5, 2020. A team of 30 trained medical students manually abstracted 274 data fields from EHRs into a REDCap database.
The study utilized Cox proportional hazards analysis to predict outcomes like death and intubation. To ensure data quality, a "double-check" system was employed for records with missing or inconsistent data, achieving high inter-rater reliability (Cohen’s kappa of 0.92–0.94). Clinical management followed specific institutional protocols for testing, admission, and respiratory support.
Results
Clinical Characteristics and Comorbidities
The median age of the cohort was 63 years, with a notable male predominance (59.6% overall, rising to 66.9% in the ICU). Only 8.2% of patients had no major comorbidities. The most prevalent baseline conditions were hypertension (60.1%), obesity (48.3%), and diabetes (37.2%).
Bimodal Intubation Distribution
A striking finding was the bimodal distribution of time from symptom onset to intubation, with peaks at 3–4 days and 9 days. This suggests a biphasic disease process where some patients deteriorate rapidly, while others remain stable for over a week before requiring mechanical ventilation. In total, 23.3% of all patients required mechanical ventilation.
Severe Renal Impact and Complications
The rate of Acute Kidney Injury (AKI) was significantly higher than in earlier international reports. Among ICU patients, 78% developed AKI, and 35.2% required dialysis. Additionally, 35.2% of the total cohort developed Acute Respiratory Distress Syndrome (ARDS).
Outcomes and Resource Utilization
ICU patients had a median length of stay of 23 days, highlighting the significant burden on hospital capacity. As of April 30, 2020, 21.1% of the total cohort had died in the hospital. Among ICU patients, the mortality rate was 43.6%, though 36.9% remained hospitalized at the time of analysis.
Discussion
This study underscores the high morbidity and mortality faced by hospitalized COVID-19 patients in NYC during the initial surge. The bimodal intubation peak and high rates of renal failure are particularly noteworthy and suggest complex underlying biological mechanisms.
Limitations
As a single-center retrospective study in an urban academic medical center, the findings may not be fully generalizable to community hospitals or other regions. Furthermore, the reliance on EHR documentation during a period of extreme clinical stress may introduce missing data. Finally, 9% of the cohort remained hospitalized at the time of publication, censoring their final outcomes.
Future research should investigate the biological drivers of the bimodal intubation distribution and the high incidence of renal failure to better tailor clinical interventions.
Summary updated