Risk Analysis Index Versus mFI-5 for Predicting Outcomes After Uvulopalatopharyngoplasty.

Abstract

Current preoperative risk stratification for uvulopalatopharyngoplasty (UPPP) primarily relies on anatomic and disease-severity measures, which do not capture physiologic reserve or vulnerability to postoperative complications. This study aimed to compare the predictive performance of the risk analysis index (RAI) and the modified frailty index-5 (mFI-5) in stratifying postoperative risk among patients undergoing UPPP. Adult patients who underwent UPPP between 2005 and 2020 were identified. Frailty was assessed, and associations with postoperative complications, length of stay, discharge disposition, and mortality were evaluated. Multivariable regression and receiver operating characteristic analysis were conducted. A total of 2129 patients were included and severe frailty (RAI > 31) was strongly associated with adverse postoperative outcomes, including Clavien-Dindo (CD) II complications (OR 21.3, 95% CI 3.7-123.8), CD IV (OR 4.6, 95% CI 2.0-10.6), extended length of stay (eLOS) (OR 29.4, 95% CI 5.4-160), non-home discharge (NHD) (OR 6.1, 95% CI 3.0-12.6), and deep surgical site infection (DSSI) (OR 17.5, 95% CI 1.7-176.3) (all p < 0.05). The mFI-5 predicted CD II, eLOS, and NHD but not CD IV or DSSI among severely frail patients (mFI-5 > 3). RAI consistently outperformed mFI-5 across outcomes, including mortality (AUC 0.813 vs. 0.580), CD II (0.784 vs. 0.652), CD IV (0.704 vs. 0.654), eLOS (0.767 vs. 0.669), and NHD (0.767 vs. 0.618). Although both RAI and mFI-5 are associated with postoperative outcomes following UPPP, RAI provides superior discrimination for mortality and major morbidity. These findings support the incorporation of RAI into preoperative risk stratification for sleep surgery to better inform perioperative planning and resource allocation.