Hospital teaching status and COVID-19 outcomes: a multicenter retrospective cohort study in the emergency department

Differences in COVID-19 outcomes may reflect variation in clinical expertise, institutional resources, and care coordination between academic teaching and non-teaching hospitals. We evaluated differences in patient characteristics, resource utilization, and mortality between hospital types during the early phase of the pandemic.

We conducted a retrospective cohort study of 2,767 adults (≥18 years) with COVID-19 presenting to five hospitals within the Johns Hopkins Health System between March 1 and 4 May 2020. Patient characteristics, clinical presentation, laboratory findings, and outcomes were compared between academic teaching and non-teaching hospitals. Group comparisons were performed using unpaired t-tests and chi-square tests. Multivariable logistic regression assessed the association between hospital teaching status and in-hospital mortality. A primary model included variables with minimal missing values, while a secondary model incorporated laboratory variables using complete-case analysis. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and the Hosmer-Lemeshow goodness-of-fit test.

Of 2,767 patients, 801 (29%) were evaluated at academic teaching hospitals. Patients at academic centers had higher comorbidity burden and more severe clinical presentation, whereas patients at non-teaching hospitals had lower access to healthcare resources. Emergency department length of stay (430 vs 269 minutes, p < 0.001), hospital length of stay (12.2 vs 4.4 days), and hospitalization costs ($55,868 vs $5,543, p < 0.001) were higher in teaching hospitals. Mortality was higher in non-teaching hospitals across both adjusted models. In the primary (n = 2,605) model (OR 6.86; 95% CI 5.18-9.10; AUC 0.70) and the secondary (n = 372) model (OR 13.4; 95% CI 7.28-24.7; AUC 0.82).

These findings suggest that structural differences, including resource availability and healthcare access, may contribute to disparities in COVID-19 outcomes. Non-teaching hospitals, while managing high volumes with limited resources, had markedly higher mortality despite lower comorbidity burden. Further investigation of system-level determinants of outcomes during public health emergencies is warranted.