RAI Outperforms mFI-5 in Predicting Post-UPPP Complications: A Frailty Stratification Breakthrough

Background

Uvulopalatopharyngoplasty (UPPP) remains a cornerstone surgical intervention for obstructive sleep apnea (OSA), yet current preoperative risk stratification relies heavily on anatomic and disease-severity measures. These fail to account for physiologic vulnerability, leaving clinicians without robust tools to predict postoperative complications. This study addresses a critical gap by comparing two frailty metrics—the Risk Analysis Index (RAI) and modified frailty index-5 (mFI-5)—in stratifying risk among UPPP patients.

Study Design

The retrospective cohort analysis included 2,129 adults who underwent UPPP between 2005 and 2020. Frailty was assessed preoperatively using RAI (a validated 14-item tool evaluating comorbidities, functional status, and nutrition) and mFI-5 (a 5-item comorbidity-based index). Outcomes included Clavien-Dindo complications (CD II/IV), extended length of stay (eLOS), non-home discharge (NHD), and mortality. Multivariable logistic regression and receiver operating characteristic (ROC) analyses determined predictive performance.

Key Findings

Frailty and Adverse Outcomes

Severe frailty (RAI >31) demonstrated striking associations: 21-fold higher odds of CD II complications (OR 21.3, 95% CI 3.7-123.8), 4.6-fold increase in CD IV events (OR 4.6, 2.0-10.6), and 17.5-fold elevated deep surgical site infection risk (OR 17.5, 1.7-176.3). mFI-5 >3 predicted CD II, eLOS, and NHD but failed to capture CD IV or DSSI.

Discriminatory Performance

RAI consistently surpassed mFI-5, with superior area-under-curve values for mortality (0.813 vs. 0.580), CD II (0.784 vs. 0.652), and eLOS (0.767 vs. 0.669). For NHD, RAI’s AUC was 0.767 versus mFI-5’s 0.618.

Expert Commentary

“This study pivots the conversation from anatomic to physiologic risk assessment,” notes Dr. A. Khan (uninvolved otolaryngologist). However, limitations include retrospective design and potential selection bias. The findings may not generalize to outpatient UPPP or pediatric populations.

Conclusion

RAI’s superior predictive validity supports its integration into preoperative workflows, potentially reducing complications through targeted resource allocation. Future research should validate these findings prospectively and explore RAI’s utility in other sleep surgeries.

Funding and Registration

Study funding was undisclosed. No clinical trial registration was reported (retrospective design).

References

1. Warrier A et al. The Laryngoscope. 2026; PMID: 41989010
2. Rockwood K et al. CMAJ. 2005; Frailty indices literature
3. Velanovich V. J Am Coll Surg. 2020; RAI validation studies