Predicting long-term mortality in older patients after non-ST-segment elevation myocardial infarction: The CRUSADE long-term mortality model and risk score.
Am Heart J. 2011 Nov;162(5):875-883.e1
Authors: Roe MT, Chen AY, Thomas L, Wang TY, Alexander KP, Hammill BG, Gibler WB, Ohman EM, Peterson ED
Abstract
OBJECTIVES: We sought to develop a long-term mortality risk prediction model and a simplified risk score for use in older patients with non-ST-segment elevation myocardial infarction (NSTEMI).
BACKGROUND: Limited data are available regarding long-term mortality rates and concomitant risk predictors after acute myocardial infarction in contemporary community practice.
METHODS: From the CRUSADE registry, a total of 43,239 (NSTEMI) patients aged ?65 years treated at 448 hospitals in the United States from 2003 to 2006 were linked to Centers for Medicare and Medicaid Services data to track longitudinal all-cause mortality (median follow-up 453 days). Cox proportional hazard modeling was used to determine baseline independent demographic, clinical, and laboratory variables associated with long-term mortality. A simplified long-term mortality risk score was subsequently developed from these results.
RESULTS: The median age of this population was 77 years, and mortality rates at 1, 2, and 3 years were 24.4%, 33.2%, and 40.3%, respectively. We identified 22 variables independently associated with long-term mortality in a full model (c-statistic 0.754 in the derivation sample and 0.744 in the validation sample). The CRUSADE long-term mortality risk score was limited to the 13 most clinically and statistically significant variables from the full model yet retained comparable discrimination in the derivation and validation samples (c-statistics 0.734 and 0.727, respectively) and had good calibration across the risk spectra.
CONCLUSIONS: Older patients face substantial long-term mortality risks after NSTEMI that can be accurately predicted from baseline characteristics. These prognostic estimates may support informed treatment decision-making and comparison of long-term provider outcomes.
PMID: 22093204 [PubMed - in process]