Percutaneous coronary intervention outcomes in US hospitals with varying structural characteristics: analysis of the NCDR®.

BACKGROUND: In the United States, there continues to be debate about whether certain types of hospitals deliver improved patient outcomes. We sought to assess the association between hospital organizational characteristics and in-hospital outcomes for percutaneous coronary intervention (PCI). METHODS: Retrospective analysis of 2004 to 2007 data for 694 US hospitals participating in the CathPCI Registry(®). Our analysis focused on 1,113,554 patients who underwent PCI in 471 not-for-profit (NFP) hospitals, 131 major teaching hospitals, 79 for-profit (FP) hospitals, and 13 physician-owned specialty hospitals. Outcomes included in-hospital mortality, stroke, bleeding, vascular injury, and a composite representing one or more of the individual complications. We used the current CathPCI Registry mortality risk model to calculate risk-standardized mortality ratios (RSMRs) for each category of hospital and compared hospital groupings for all patients in aggregate and in subgroups stratified by patients' indications for PCI. RESULTS: Patients treated in major teaching hospitals were younger, whereas FP hospitals performed a greater proportion of PCI for patients with ST-elevation myocardial infarction (P < .0001). Specialty hospitals treated patients with less acuity, including a lower proportion of patients with ST-elevation myocardial infarction. In unadjusted analyses, specialty hospitals had significantly lower rates of all adverse outcomes compared with NFP, teaching, and FP hospitals including in-hospital mortality (0.7%, 1.2%, 1.4%, and 1.4%, respectively; P < .001) and the composite end point (2.4%, 4.1%, 4.6%, and 4.3%, respectively; P < .001). In adjusted analyses, RSMR was significantly lower for specialty hospitals when compared with the other 3 groups for all patients in aggregate (RSMR 1.05%, 1.30%, 1.38%, 1.39%; P < .001); these differences remained clinically significant but were no longer statistically significant in subgroup analyses. CONCLUSIONS: Specialty hospitals appear to have lower rates of most adverse outcomes for PCI. Specialty hospitals may have developed expertise in narrow procedural areas that could be adapted to the larger population of general hospitals.