Healthcare-Associated COVID-19 in Ontario, Canada: Relative Mortality and Contribution to Community Epidemic Growth
Wilson, N. J.; Grima, A.; Lee, E. C.; Fisman, D.
Show abstract
BackgroundThe COVID-19 pandemic placed immense strain on Canadas healthcare system and disproportionately affected individuals with poorer baseline health. Healthcare-associated infections (HAIs) increase risk for both patients and healthcare workers and are often more severe due to advanced age and comorbidities. While efforts have aimed to reduce in-hospital transmission, the individual- and community-level consequences of HAIs require further study. We aimed to assess whether healthcare-associated COVID-19 cases had higher odds of death compared to hospitalized community-acquired cases, and to evaluate the directionality of transmission between hospitals and the community. MethodsWe analyzed COVID-19 surveillance data from Ontarios Case Contact and Management System and the COVaxON vaccine registry (March 17, 2020, to September 4, 2022). Latent class analysis was used to classify hospitalized cases by likelihood of healthcare-associated infection. Mortality odds by category were estimated using binomial logistic regression. Directionality between hospital outbreaks and community cases was assessed using a modified Granger causality approach. FindingsCompared to patients with low likelihood of healthcare-associated infection, those moderately likely to have acquired COVID-19 in hospital had elevated odds of death (OR: 1.26, 95% CI: 1.14-1.40); no significant increase was seen in the high-likelihood group (OR: 1.05, 95% CI: 0.96-1.15). Community cases did not predict hospital outbreaks (p=0.5749), but hospital outbreaks predicted community case growth (p<0.0001). InterpretationHospital-acquired COVID-19 is associated with excess mortality and may drive community transmission. Preventing in-hospital transmission is critical to protecting patients and controlling broader epidemic spread. FundingSupported by a Canadian Institutes for Health Research project grant, #518192.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.