作者: Stephen E. Kimmel , Jesse A. Berlin , Brian L. Strom , Warren K. Laskey
DOI: 10.1016/0735-1097(95)00294-4
关键词: Stepwise regression 、 Internal medicine 、 Cardiology 、 Logistic regression 、 Angioplasty 、 Medicine 、 Risk assessment 、 Myocardial infarction 、 Bypass surgery 、 Risk factor 、 Unstable angina
摘要: Objectives. This study was designed to determine the preprocedural risk factors for major complications (emergent coronary bypass surgery, myocardial infarction or death) of angioplasty and derive validate a simplified index that predicts patients' priori complications. Background. Previous studies after may not be generalizable current, broad-based practice. Furthermore, our knowledge clinically useful predictive has been derived independently validated. Methods. From data collected prospectively Registry Society Cardiac Angiography Interventions 1992, multivariable logistic regression used which variables were associated with in 10,622 first procedures. Stepwise receiver operating characteristic curves then this registry develop validated using 5,250 procedures 1993 registry. Results. Predictors multivessel disease, unstable angina, recent infarction, type C lesion left main angioplasty, shock, age, geographic region absence previous surgery. The consisted six these plus aortic valve disease classified patients into four groups: low (1.3% complications), moderate (2.8%), high (12.7%) very (29.7%) risk. demonstrated consistent reliability discriminatory ability when applied data. Conclusions. identified selected populations also apply currently variables, can stratify groups before thus aiding assessment, resource allocation adjustment.