Results:

The median age for all patients was 68 years, 68% were men, and 17% had had a previous MI. Overall, 309 patients (10%) presented with a Killip class III or higher and the median predicted risk for in-hospital mortality was 5.8%. Patients admitted to academic hospitals were younger, had lower predicted risk for in-hospital mortality, and were more likely mobile intensive care unit users. More patients admitted to nonacademic hospitals had previous congestive heart failure. Cardiac arrest on admission (P for interaction = 0.006) and previous stroke (P for interaction = 0.04) accounted for a higher proportion of admissions to academic hospitals among patients with non-STEMI only. The percentages of patients with STEMI receiving acute reperfusion therapy (i.e., thrombolysis or percutaneous coronary intervention [PCI]) were 68%, 65%, and 36% for academic hospitals, nonacademic hospitals with PCI capability, and nonacademic hospitals without PCI capability. Unadjusted 1-year mortality for academic and nonacademic hospitals was 10% versus 15% for patients with STEMI (p = 0.01) and 13% versus 14% for patients with non-STEMI (p = 0.75). Patients treated in academic or nonacademic hospitals with PCI capability were more likely to receive reperfusion and recommended drug therapies than those treated in nonacademic hospitals without PCI. After adjusting for baseline characteristics, the hazards of death associated with admission to nonacademic hospitals with and without PCI capability relative to academic hospitals were 1.13 (95% CI, 0.79-1.62) and 1.65 (95% CI, 1.09-2.49) for those with STEMI and 0.95 (95% CI, 0.66-1.36) and 1.06 (95% CI, 0.72-1.58) for those with non-STEMI. Further adjustment for receipt of acute reperfusion and recommended drug therapies eliminated all differences in mortality between the study groups.