Abstract
BACKGROUND: Little is known about how differences in out of hospital cardiac arrest patient volume affect in-hospital myocardial infarction (MI) mortality. HYPOTHESIS: Hospitals accepting cardiac arrest transfers will have increased hospital MI mortality. METHODS: MI patients (ST elevation MI [STEMI] and non-ST elevation MI [NSTEMI]) in the Acute Coronary Treatment Intervention Outcomes Network Registry were included. Hospital variation of cardiac arrest and temporal trend of the proportion of cardiac arrest MI patients were explored. Hospitals were divided into tertiles based on the proportion of cardiac arrest MI patients, and association between in-hospital mortality and hospital tertiles of cardiac arrest was compared using logistic regression adjusting for case mix. RESULTS: A total of 252 882 patients from 224 hospitals were included, of whom 9682 (3.8%) had cardiac arrest (1.6% of NSTEMI and 7.5% of STEMI patients). The proportion of MI patients who had cardiac arrest admitted to each hospital was relatively low (median 3.7% [25th, 75th percentiles: 3.0%, 4.5%]).with a range of 4.2% to 12.4% in the high-volume tertiles. Unadjusted in-hospital mortality increased with tertile: low 3.8%, intermediate 4.6%, and high 4.7% (P < 0.001); this was no longer significantly different after adjustment (intermediate vs high tertile odds ratio (OR) = 1.02; 95% confidence interval [0.90-1.16], low vs high tertile OR = 0.93 [0.83, 1.05]). CONCLUSIONS: The proportion of MI patients who have cardiac arrest is low. In-hospital mortality among all MI patients did not differ significantly between hospitals that had increased proportions of cardiac arrest MI patients. For most hospitals, overall MI mortality is unlikely to be adversely affected by treating cardiac arrest patients with MI.