Location and outcomes of rehospitalizations after critical illness in a single-payer healthcare system.

Abstract

Unplanned rehospitalization at a hospital other than the initial hospital may contribute to poor outcomes. We examined the location of rehospitalizations and assessed outcomes following critical illness in a single-payer healthcare system. Population-based retrospective cohort study using linked datasets (2012-2017) from Ontario, Canada including adults (≥18 years) with an unplanned rehospitalization within 30-days after an index hospitalization that included an ICU stay with mechanical ventilation. Outcomes were the percentage of 30-day rehospitalizations at non-index hospitals, mortality and costs. We employed logistic regression and generalized linear models to assess associations. There were 14,997 (16.4%) 30-day rehospitalizations. Of these 2765 (18.4%) occurred in a non-index hospital. Distance of home residence from the index hospital was the strongest predictor of a non-index rehospitalization (adjusted odds ratio (aOR) 8.40, 95%CI 7.05-10.01, highest vs. lowest distance quintile). Within 30-days of rehospitalization, deaths (aOR 0.91, 95%CI (0.80-1.04)) and total healthcare costs (adjusted relative risk 1.03 (1.00-1.06)), were similar for patients readmitted to the index or a non-index hospital. Non-index rehospitalization within 30-days of initial discharge is common following critical illness. These rehospitalizations were not significantly associated with an increased risk of harm or higher costs in a single-payer healthcare system.