Abstract
BACKGROUND: In previous studies, there was an increase in mortality with secondary coinfections in all COVID-19 variants. However, no prior study has explored the association of coinfection with outcomes of hospitalized patients among the COVID-19 variants (Alpha, Delta, and Omicron). METHODS: This observational cohort study involved 21,186 patients hospitalized with COVID-19 in 25 hospitals in Texas. Patients were divided into groups by surges of COVID-19: Alpha (November 1, 2020-February 10, 2021), Delta (July 10, 2021-October 14, 2021), and Omicron (December 21, 2021-March 3, 2022). Data were collected from electronic health records using methodology from the Viral Respiratory Illness Universal Study COVID-19 registry (NCT04323787) of COVID-19 hospitalizations. Multivariable Cox-proportional hazard regression model assessed the adjusted effect of different surge periods on mortality. RESULTS: Bacterial coinfections varied among hospitalization surges associated with Alpha (8.5%), Delta (11.7%), and Omicron (11.9%) variants. Adjusted analyses showed a higher 30-day and 90-day mortality in all variants when coinfections were present compared with isolated COVID-19 infection. In particular, 30-day and 90-day mortality were significantly worse with Delta compared to Alpha and Omicron. CONCLUSIONS: All variants were associated with a higher mortality when bacterial coinfections were present. Delta was associated with a higher risk-adjusted mortality at 30 days and thereafter.