Preprocedural Risk Factors of In‐Hospital Mortality Following Percutaneous Coronary Intervention: A Systematic Review of Risk‐Adjustment Models

Assessing the preprocedural risk of mortality following percutaneous coronary intervention (PCI) is crucial for clinical decision‐making, quality registries, and performance monitoring. This systematic review aims to summarize preprocedural factors associated with in‐hospital mortality post‐PCI. A systematic search of MEDLINE, EMBASE, CINAHL, and Web of Science was conducted up to April 2024, without language restrictions, to identify preprocedural factors associated with in‐hospital mortality post‐PCI. The information was systematically evaluated and descriptively summarized following the CHhecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modeling Studies (CHARMS) checklist. The search initially identified 9552 studies, with 28 articles ultimately included in the final analysis. The pooled prevalence of in‐hospital mortality post‐PCI was 1.63%, whereas it was 6.40% for patients with acute coronary syndrome (ACS). A total of 77 independent preprocedural factors were found to be significantly associated with in‐hospital mortality, of which 22 were consistently reported across multiple studies and identified as potential key factors. Factors most frequently reported in various studies included, but not limited to, ACS ( n = 27, 96.4%), renal failure ( n = 24, 85.7%), age ( n = 19, 67.9%), gender ( n = 16, 57.1%), cardiac events ( n = 15, 53.6%), ejection fraction ( n = 15, 53.6%), vascular disease ( n = 12, 42.9%), heart failure ( n = 12, 42.9%), body mass index ( n = 10, 35.7%), lesion location ( n = 9, 32.1%), diabetes ( n = 8, 28.6%), and vessel disease ( n = 8, 28.6%). Eleven articles (39.2%) used imputation methods to handle missing data. Logistic regression was used by 19 articles (67.9%), and five articles (18.6%) used the machine learning (ML) method. Nineteen articles (67.9%) reported the discriminatory performance of the models using receiver operating characteristic (ROC) score ranging from 0.776 to 0.960. Incorporating the 22 preprocedural factors identified in this study can aid clinicians in decision‐making for high‐risk cases. To improve risk adjustment models, these factors should be integrated, missing data carefully managed, models validated, and ML techniques utilized.