Between-hospital variability in outcomes after cardiac arrest in a large clinical trial network.

Abstract

Previous cohorts found significant variability in outcomes of out-of-hospital cardiac arrest (OHCA) between hospitals. We analyzed screening logs of a recent multicenter trial to quantify between-hospital variability in risk-adjusted OHCA survival. Among 69 hospitals participating in the Influence of Cooling Duration on Efficacy in Cardiac Arrest Patients (ICECAP) trial, we extracted data pertaining to both screen failures and enrollments between April 2020 and March 2024. We included all patients screened for ICECAP after resuscitation from OHCA, regardless of whether they were subsequently enrolled, excluding those who died prior to admission, those aged <18 years, those awake on presentation, and those with severe preexisting functional impairment or terminal illness. We summarized population characteristics and hospital-level outcomes, then quantified between-hospital variability in survival to discharge overall and after adjustment for patient factors using Bayesian hierarchical regressions. In secondary analyses, we explored relationships between site volume, and temperature management practices and risk-adjusted survival. Overall, 8603 patients were included of whom 2192 (25%) survived to discharge. Median survival to discharge by site was 28% [IQR 24-32%], equating to a median odds ratio (MOR) of 1.80 (95% credible interval [CI] 1.72-1.91) across sites. After adjusting for measured patient characteristics, between-site variability increased slightly (MOR 1.90 [95% CI 1.80-2.03]). There was no relationship between site volume or temperature management practices and risk-adjusted outcomes. This prospective, multicenter study suggests that the marked between-hospital variability in OHCA outcomes consistently observed in previous cohorts persists today. Improving outcomes at lower performance hospitals could have a substantial impact on survival and functional recovery after resuscitation from cardiac arrest.