Cardiopulmonary resuscitation outcomes in hospitalized community-dwelling individuals and nursing home residents based on activities of daily living

Objectives: To determine whether poor functional status is associated with worse outcomes after attempted cardiopulmonary resuscitation (CPR). Design: Retrospective study of individuals who experienced cardiac arrest stratified according to dependence in activities of daily living (ADLs) and residential status (nursing home (NH) or community dwelling). Setting: Two hundred thirty-five hospitals throughout North America. Participants Adult inpatients aged 65 and older who had experienced a cardiac arrest as reported to the Get with the Guidelines - Resuscitation registry between 2000 and 2008. Measurements: Primary outcomes were return of spontaneous circulation (ROSC) and survival to discharge. Results: Twenty-six thousand three hundred twenty-nine individuals who experienced cardiac arrest met inclusion criteria. NH residents dependent in ADLs had a lower odds than community-dwelling independent participants of achieving ROSC (odds ratio (OR) = 0.73, 95% confidence interval (CI) = 0.63-0.85), whereas participants dependent in ADLs from either residential setting had lower odds of survival (community-dwelling: OR = 0.76, 95% CI = 0.63-0.92; NH: OR = 0.79, 95% CI = 0.64-0.96) after adjusting for participant and arrest characteristics. Duration of resuscitation and doses of epinephrine or vasopressin were similar between groups and had no significant effect on ROSC or survival, although participants dependent in ADLs were more likely to have a do-not-resuscitate (DNR) order placed after ROSC. Overall, median time to signing a DNR order after resuscitation was 10 hours (interquartile range 2-70). Conclusion: Functional and residential status are important predictors of survival after in-hospital cardiac arrest. Contrary to the hypothesis but reassuring from a quality-of-care perspective, less-aggressive attempts at resuscitation do not appear to contribute to poorer outcomes in individuals dependent in ADL, regardless of residential status.