Updated CRUSADE score to predict in‐hospital bleeding: External validation in the Thai percutaneous coronary intervention registry

Abstract

AbstractBackgroundThe Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes with Early Implementation of the ACC/AHA Guidelines (CRUSADE) score has been recommended to predict in‐hospital bleeding risk in non‐ST segment elevation myocardial infarction (NSTEMI) patients. The evaluation of the CRUSADE risk score in Asian patients undergoing contemporary percutaneous coronary intervention (PCI) for NSTEMI is necessary.AimsWe aimed to validate and update the CRUSADE score to predict in‐hospital major bleeding in NSTEMI patients treated with PCI.MethodThe Thai PCI registry is a large, prospective, multicenter PCI registry in Thailand enrolling patients between May 2018 and August 2019. The CRUSADE score was calculated based on 8 predictors including sex, diabetes, prior vascular disease (PVD), congestive heart failure (CHF), creatinine clearance (CrCl), hematocrit, systolic blood pressure, and heart rate (HR). The score was fitted to in‐hospital major bleeding using the logistic regression. The original score was revised and updated for simplification.ResultsOf 19,701 patients in the Thai PCI registry, 5976 patients presented with NSTEMI. The CRUSADE score was calculated in 5882 patients who had all variables of the score available. Thirty‐five percent were female, with a median age of 65.1 years. The proportion of diabetes, PVD, and CHF was 46%, 7.9%, and 11.2%, respectively. The original and revised models of the CRUSADE risk score had C‐statistics of 0.817 (95% CI: 0.762–0.871) and 0.839 (95% CI: 0.789–0.889) respectively. The simplified CRUSADE score which contained only four variables (hematocrit, CrCl, HR, and CHF), had C‐statistics of 0.837 (0.787–0.886). The calibration of the recalibrated, revised, and simplified model was optimal.ConclusionsThe full and simplified CRUSADE scores performed well in NSTEMI treated with PCI in Thai population.