The association of duration of resuscitation and long-term survival and functional outcomes after out-of-hospital cardiac arrest

Longer emergency medical system cardiopulmonary-resuscitation-to-return of-spontaneous-circulation (EMS CPR-to-ROSC) interval has been associated with worse hospital discharge outcomes after out-of-hospital cardiac arrest (OHCA). We hypothesized that this association extends post-discharge in hospital survivors. We investigated whether pre-arrest co-morbidities influence the duration of resuscitation.We included EMS-treated adult OHCA (January 2009 - December 2016) from British Columbia Cardiac Arrest Registry linked to provincial databases. Pre-OHCA characteristics were compared by ≤10, 10-20, and >20 min interval categories. Outcomes included survival and functional outcomes at hospital discharge and 1- and 3-year survival. We examined the relationship between CPR-to-ROSC intervals and survival using Kaplan-Meier. We examined the relationship between the CPR-to-ROSC interval (continuous variable) with all outcomes using regression models.Among 10,241 OHCA, 4604 (45%) achieved ROSC, with a median CPR-to-ROSC interval of 15.5 (IQR 9.0-22.9) minutes. Diabetes, chronic kidney disease, and prior myocardial infarction were associated with longer CPR-to-ROSC intervals. 1245 (12.2%) survived to hospital discharge. Among hospital survivors, Kaplan-Meier survival at 1- and 3-years were 92% [95% CI 90-93%] and 84% [95% CI 82-86%] respectively; survival curves stratified by CPR-to-ROSC intervals were not statistically different. Longer CPR-to-ROSC interval was non-linearly associated with lower survival and functional outcomes at hospital discharge but not with post-discharge outcomes.Longer CPR-to-ROSC interval was associated with lower survival at hospital discharge and was influenced by pre-arrest co-morbidities. However, these intervals were not associated with long-term survival or functional outcome among hospital survivors, suggesting early risk of longer CPR-to-ROSC intervals does not persist.