Derivation and validation of a combined in-hospital mortality and bleeding risk model in acute myocardial infarction

Abstract Background In the potent new antiplatelet era, it is important issue how to balance the ischemic risk and the bleeding risk. However, previous risk models have been developed separately for in-hospital mortality and major bleeding risk. Therefore, we aimed to develop and validate a novel combined model to predict the combined risk of in-hospital mortality and major bleeding at the same time for initial decision making in patients with acute myocardial infarction (AMI). Methods Variables from the Korean Acute Myocardial Infarction Registry (KAMIR) – National Institute of Health (NIH) database were used to derive (n = 8955) and validate (n = 3838) a multivariate logistic regression model. Major adverse cardiovascular events (MACEs) were defined as in-hospital death and major bleeding. Results Seven factors were associated with MACE in the model: age, Killip class, systolic blood pressure, heart rate, serum glucose, glomerular filtration rate, and initial diagnosis. The risk model discriminated well in the derivation (c-static = 0.80) and validation (c-static = 0.80) cohorts. The KAMIR-NIH risk score was developed from the model and corresponded well with observed MACEs: very low risk (0.9%), low risk (1.7%), moderate risk (4.2%), high risk (8.6%), and very high risk (23.3%). In patients with MACEs, a KAMIR-NIH risk score ≤ 10 was associated with high bleeding risk, whereas a KAMIR-NIH risk score > 10 was associated with high in-hospital mortality. Conclusion The KAMIR-NIH in-hospital MACEs model using baseline variables stratifies comprehensive risk for in-hospital mortality and major bleeding, and is useful for guiding initial decision making.