Nationwide 30-Day Surgical Outcomes in Korea: An Interim Analysis of the Korean Quality Improvement Platform in Surgery

Abstract

BACKGROUND:National surgical registries such as the American College of Surgeons NSQIP and Japan's National Clinical Database have shown that structured, risk-adjusted outcome monitoring improves surgical quality. However, no comparable nationwide registry has been reported in Korea.STUDY DESIGN:This interim analysis of the Korean Quality Improvement Platform in Surgery included 42,943 cases from 45 hospitals (2013 to 2023), covering gastric, colorectal, hepatic, pancreatic surgery, and kidney transplantation. Outcomes were 30-day complications, major complications (Clavien-Dindo grade III or higher), and mortality. Hierarchical generalized linear mixed models generated risk-adjusted estimates, accounting for patient- and procedure-level factors and stabilizing results for low-volume centers.RESULTS:The 30-day complication, major complication, and mortality rates were 38.7%, 12.7%, and 0.59%, respectively. Mortality was highest after pancreatic surgery (1.17%) and lowest after gastric surgery (0.16%). Pneumonia occurred in 1.74% of cases and surgical site infection (SSI) in 4.0%. Risk-adjusted benchmarking of a representative institution (H037) classified pneumonia (adjusted odds ratio [OR] 0.67) as "as expected," whereas SSI (adjusted OR 0.15) and mortality (adjusted OR 0.55) were classified as "exemplary." Multivariable analyses confirmed established risk factors: colorectal surgery and higher wound class predicted SSI, chronic pulmonary disease and emergency surgery predicted pneumonia, and higher American Society of Anesthesiologists class and age predicted mortality.CONCLUSIONS:Korean Quality Improvement Platform in Surgery demonstrates the feasibility of a nationwide, risk-adjusted surgical registry in Korea and establishes a benchmark for institutional feedback. As data collection nears completion (target: 219,200 cases), future efforts will focus on subspecialty-specific models, patient-level risk calculators, and integration with artificial intelligence-driven clinical decision-support systems.