Simple risk-score model for in-hospital major bleeding based on multiple blood variables in patients with acute myocardial infarction

Background: In-hospital bleeding is associated with poor prognosis in patients with acute myocardial infarction (AMI). We sought to investigate whether a combination of pre-procedural blood tests could predict the incidence of in-hospital major bleeding in patients with AMI. Methods and results: A total of 1684 consecutive AMI patients who underwent primary percutaneous coronary intervention (PCI) were recruited and randomly divided into derivation (n = 1010) and validation (n = 674) cohorts. A risk-score model was created based on a combination of parameters assessed on routine blood tests on admission. In the derivation cohort, multivariate analysis revealed that the following 5 variables were significantly associated with in-hospital major bleeding: hemoglobin level < 12 g/dL (odds ratio [OR], 3.32), white blood cell count >10,000/mu L (OR, 2.58), platelet count <150,000/mu L (OR, 2.51), albumin level < 3.8 mg/dL (OR, 2.51), and estimated glomerular filtration rate < 60 mL/min/1.73 m2 (OR, 2.31). Zero to five points were given according to the number of these factors each patient had. Incremental risk scores were significantly associated with a higher incidence of in-hospital major bleeding in both cohorts (P < 0.001). Receiver operating characteristic curve analysis of risk models showed adequate discrimination between patients with and without in-hospital major bleeding (derivation cohort: area under the curve [AUC], 0.807; 95% confidence interval [CI], 0.759-0.848; validation cohort: AUC, 0.793; 95% CI, 0.725-0.847). Conclusions: Our novel laboratory-based bleeding risk model could be useful for simple and objective prediction of in-hospital major bleeding events in patients with AMI.