To compare the predictive performance of the RAI versus the mFI-5 for postoperative outcomes following ASD surgery.

Adult spinal deformity (ASD) surgery is associated with complication rates of 30–70%. Accurate preoperative risk stratification is critical, yet the commonly used modified Frailty Index-5 (mFI-5) may not optimally predict adverse outcomes. The Risk Analysis Index (RAI), a validated frailty measure, may provide superior predictive performance.

We performed a retrospective cohort analysis of ACS-NSQIP data (2015–2021) including 632 patients undergoing ASD surgery (CPT 22800, 22802, 22804). Frailty was assessed using RAI and mFI-5. Primary outcomes were mortality and non-home discharge, defined a priori as outcomes where RAI achieved AUC >0.7 and significantly outperformed mFI-5. Secondary outcomes included complications, readmissions, reoperations, and extended length of stay. Model discrimination was compared using AUROC analysis with bootstrap validation.

RAI demonstrated superior discrimination for mortality (AUC 0.981 vs mFI-5 0.690, p<0.001) and non-home discharge (0.787 vs 0.622, p<0.001). Non-home discharge rates increased across RAI categories: Robust 18.8%, Normal 63.5%, Frail 66.7%, Very Frail 100% (p<0.001). For secondary outcomes, RAI showed modest improvements over mFI-5 for extended length of stay (0.629 vs 0.571) and major complications (0.593 vs 0.550). Bootstrap validation confirmed model stability.

The RAI significantly outperformed mFI-5 in predicting mortality and discharge disposition following ASD surgery. These findings support the integration of RAI into preoperative assessment to optimize surgical decision-making, counseling, and resource allocation in this high-risk population.