Thank you for your interest in CCO content. As a guest, please complete the following information fields. These data help ensure our continued delivery of impactful education.
Become a member (or login)? Member benefits include accreditation certificates, downloadable slides, and decision support tools.
Assistant Professor of Medicine
Division of Pulmonary, Critical Care, & Sleep Medicine
New York University School of Medicine
Director, Medical Intensive Care Unit
New York, New York
Vikramjit Mukherjee, MD, has no relevant conflicts of interest to report.
Recently, I discussed data on racial disparities in COVID-19 outcomes, demonstrating that comorbidities and certain sociodemographic factors were associated with worse COVID-19 outcomes. I’d like to continue exploring the theme of nonclinical factors affecting COVID-19 outcomes by discussing new data on the effects of COVID-19 caseload on patient mortality.
The 11 acute care hospitals of the New York City Health + Hospitals Network (NYC H+H) were at the epicenter of the COVID-19 pandemic in the United States in March, April, and May 2020. During this time, the NYC H+H experienced a surge of critically ill patients who required ICU support for extended periods of time. We recently published a retrospective analysis of mortality rates in 2233 patients with confirmed COVID-19 admitted to NYC H+H from March 24 to May 12, 2020. During that time period, our network ICU directors reported their respective ICU surge levels each day, using an internally developed scale:
We found that the mortality rate in ICUs with a surge level ≥3 was significantly greater than in ICUs with a surge level <3 (68.2% vs 57.1%, respectively; P <.0001). Of note, we also saw associations between poverty levels and ICU surge levels. When overall surge levels decreased, patients from neighborhoods with medium to very high poverty levels had reduced mortality rates, whereas the mortality rate among patients in low-poverty neighborhoods remained steady, indicating that some hospitals experienced greater surge strain than others. Overall, patients experienced 1.4 times greater odds of dying if admitted to ICU-level care during periods with the highest surge levels.
A larger retrospective cohort study published by Kadri and colleagues analyzed COVID-19–related mortality rates and surge effects among 144,116 patients hospitalized with confirmed COVID-19 in 558 US hospitals from March to August 2020. These investigators created an innovative surge index that incorporated the severity-weighted COVID-19 caseload, nursing needs, and baseline hospital bed capacity.
The investigators observed that patients with COVID-19 who were admitted to a hospital with a high surge index died more frequently than patients who were admitted to a hospital with a low surge index. For example, their analysis yielded an adjusted odds ratio of mortality of 2.00 (95% CI: 1.69-2.38) in hospitals in the 99th percentile of the surge index compared with hospitals below the 50th percentile. The investigators estimated that 5868 of 25,344 total COVID-19 deaths (23.2%) were potentially attributable to COVID-19 surge effects on hospitals.
What if one in 4 COVID-19 deaths during a surge could be avoided by improving the processes that allow hospitals to share resources and shift ICU-level patients from overloaded hospitals to those with capacity? There are very few nonclinical interventions that have the potential to reduce mortality as substantially. Some, but not all, cities and states have such hospital level–loading programs in place. Going forward, these new data highlight the critical role of level-loading strategies in pandemic responsiveness.
Was your hospital strained to or beyond capacity by a COVID-19 surge? If so, how effective was the level-loading response? Answer the polling question and join the conversation by posting in the discussion section.