Abstract 11420: Long-Term Mortality, MACE, and Angina Among Patients With Diabetes and Myocardial Infarction
Background: Patients with diabetes mellitus (DM) are at higher risk for adverse events after MI, including death and recurrent ischemia. As such, we sought to develop prediction models to identify DM patients at highest risk for these outcomes.
Methods: We examined 3 adverse outcomes (3-year mortality, 1-year angina, 1-year MACE [mortality, MI, stroke, urgent revascularization]) among patients with DM discharged alive after MI from 24 US hospitals. Using Harrell’s backward selection, we developed hierarchical multivariable models with a parsimonious set of covariates for each outcome. The mortality and angina models were validated in a separate 19-site US MI registry.
Results: Among 1626 DM patients, event rates were 19% for 3-year mortality, 11% for 1-year MACE, and 27% for 1-year angina. The models had good discrimination, with c-indices of 0.80, 0.72, and 0.73, respectively, and excellent calibration. The strongest predictors (in terms of F-value) were age, hemoglobin, and creatinine for mortality; depression, hemoglobin, and history of heart failure for MACE; and angina prior to MI, age, and prior CABG for angina (Table). The observed rates of outcomes in the lowest to highest decile of predicted risk ranged broadly for all 3 outcomes, with ranges of 1-58% for mortality, 1-38% for MACE, and 8-62% for angina. The mortality and angina models performed well in the validation cohort, although they tended to over-predict mortality at higher levels of risk and under-predict angina at lower levels of risk.
Conclusion: Patients with DM and MI represent a particularly high-risk cohort, with increased mortality, MACE, and angina. However, patients with DM are not a homogenous group, as there is a broad range of risk levels for adverse outcomes within the DM population. We have developed and validated models so as to identify the highest-risk DM patients for whom closer follow-up and secondary prevention strategies can be targeted most aggressively.
Author Disclosures: S.V. Arnold: None. J.A. Spertus: Research Grant; Modest; Gilead Sciences, NHLBI, ACCF, AHA, Eli Lilly, EvaHeart, Amorcyte. Ownership Interest; Modest; SAQ, KCCQ, PAQ. Consultant/Advisory Board; Modest; Genentech, Amgen, United Healthcare. D.K. McGuire: Consultant/Advisory Board; Modest; Takeda, Janssen, Merck, Regeneron, Boehringer Ingelheim, Orexigen, AstraZeneca, Bristol Myers Squibb, Eli Lilly, Daiichi Sankyo, Eisai, Omthera, Lexicon. K.J. Lipska: None. Y. Li: None. Y. Xu: Employment; Significant; Genentech. J.M. Stolker: Research Grant; Modest; GE Healthcare. Speakers Bureau; Modest; AstraZeneca, Astellas, InfraReDx LLC. Consultant/Advisory Board; Modest; Cordis Corp. A. Goyal: None. M. Kosiborod: Research Grant; Modest; AHA, Genentech, Sanofi-Aventis, Gilead, Medtronic Minimed, Glumetrics, Maquet, Eisai. Consultant/Advisory Board; Modest; Genentech, Gilead, F Hoffmann-La Roche, Medtronic Minimed, AstraZeneca, Abbvie, Regeneron, Edwards Lifesciences, Eli Lilly.
- © 2014 by American Heart Association, Inc.