Abstract 65: Racial Disparities in Outcomes Following Nonshockable In-Hospital Cardiac Arrest

Abstract

Background: Black patients have previously been shown to have worse outcomes from ventricular arrhythmias. This discrepancy was largely explained by individual patient characteristics and racial clustering by hospital. However, the vast majority of in-hospital cardiac arrests are due to pulseless electrical activity (PEA) and asystole and little is known about the relationship between race and outcomes in this patient group. Methods: We analyzed PEA and asystolic in-hospital cardiac arrests in the Get With The Guidelines-Resuscitation (GWTG-R) database from 2000 to 2009. A generalized estimating equation model was used to estimate the average survival difference between black and white patients in the study population. Multilevel conditional fixed effects logistic regression models were then used to estimate the relationship between race and survival to discharge and ROSC, sequentially controlling for hospital, patient demographics, comorbidities, arrest characteristic, process measures and interventions in place at the time of arrest. Results: We identified 76,385 eligible index arrests with a presenting rhythm of PEA or asystole, of which 19,236 (25%) were black. Black patients were younger on average (63 vs 71 years, p<0.001) and had more comorbidities than their white counterparts. Survival to discharge was 13% for white patients and 10% for black patients (unadjusted OR 0.83; 95% CI 0.78-0.88). This effect remained nearly constant after controlling for hospital, patient, and event-specific confounders (adjusted OR 0.85; 95% CI 0.79-0.92). Similar results were shown for ROSC (adjusted OR 0.88; 95% CI 0.84-0.92). Conclusion: Black patients have significantly lower survival rates after in-hospital PEA and asystolic arrests. Hospital, patient demographics, event characteristics, and co-morbidities as measured in the GWTG-R database failed to explain this discrepancy. Furthermore, the consistency between the outcomes differences in ROSC and survival to discharge suggest that there are unmeasured intra-arrest and/or pre-arrest racial differences present. This is in contrast to shockable cardiac arrests in the same database, in which the disparity was greatly reduced after accounting for measured confounders.