Abstract 11056: Hospitals Accepting Greater Proportions of High Risk Transfer-In STEMI Patients Do Not Have Higher Risk-Adjusted Hospital Mortality: A Report From the AHA Mission: Lifeline Program
Background: Mortality is an important quality measure for acute MI hospital care. There is concern that, despite risk adjustment, PCI receiving hospitals receiving a disproportionate volume of high risk STEMI transfers may have their reported mortality adversely affected. Methods: All STEMI patients from April 2011 to December 2013 in the ACTION Registry®-GWTG™ were included. High risk was defined as pts with either cardiogenic shock or cardiac arrest on admission. Hospitals were divided into tertiles based on the proportion of high risk STEMI patients who were transferred relative to the total number of STEMI patients treated. Adjusting for covariates in the ACTION mortality risk model, the differences in risk-adjusted in-hospital mortality in each tertile were determined before and after excluding high risk STEMI transfer pts. Results: Among 119,680 STEMI pts treated at 539 primary PCI hospitals, 37,028 (31%) pts were transfers, of whom 4,500 (4%) were high risk. The proportion of high risk STEMI transfers ranged from 0-12% across hospitals. Times from initial hospital presentation to PCI were similar across tertiles: Low 107 min; Middle, 100 min; High 106 min. The ACTION mortality risk model, which includes cardiogenic shock but not cardiac arrest, slightly underestimated mortality for high-risk STEMI transfer pts (observed in-hospital mortality rate: 26%, predicted mortality rate: 24%). While differences in observed hospital mortality were present among hospitals with a greater proportion of high-risk transfers, risk-adjusted mortality was unaffected by the inclusion or exclusion of high-risk transfer patients across all tertiles (TABLE). Conclusions: Receiving PCI hospitals accepting greater proportions of high risk STEMI transfer pts did not have a higher risk-adjusted in-hospital mortality when a clinical mortality risk model was used for risk adjustment.