The revised risk analysis index outperforms the 5-factor modified frailty index in predicting postoperative morbidity...

Pilon fractures often result from high energy mechanisms resulting in long-term morbidity and postoperative complications. While there is extensive literature investigating frailty scales as tools for preoperative risk stratification in orthopedic surgery, there is no literature regarding their use in fixation of pilon fractures. The objective of this study was to compare the predictive ability and discriminative accuracy of the Revised Risk Analysis Index (RAI-rev) and the 5-Item Modified Frailty (mFI-5) in 30-day postoperative outcomes following surgical fixation of pilon fractures.

The ACS-NSQIP database was used to identify patients undergoing surgical fixation of pilon fractures from 2015 to 2020. Multivariate analysis was used to analyze the predictive ability of frailty status for 30-day postoperative outcomes. Discriminatory accuracy of frailty scales was assessed using receiver operating characteristic analysis. Significance was determined by p < 0.05.

This study included 2935 patients, aged 33-59, of which 53 % (1562/2935) were male. The RAI-rev categorized 44.2 % (1296/2935) as normal, 44.6 % (1309/2935) as frail, and 11.2 % (330/2935) as severely frail. The mFI-5 was a significant predictor of major complications, readmission, and CDIV complications for severely frail patients and of eLOS, and NHD in both frail and severely frail patients. The RAI-rev had superior discriminative accuracy for most outcomes with good discriminative ability for NHD compared to mFI-5 (0.79 vs 0.74).

This study demonstrated that compared to RAI-rev, mFI-5 is a significant predictor of postoperative complications after surgical fixation of pilon fractures, but RAI-rev had more accurate predictive ability when compared to mFI-5, most notable for NHD (AUC 0.79 vs 0.74). However, ROC values for both RAI (AUC 0.65-0.79) and mFI-5 (AUC 0.59-0.74) still fell below an AUC of 0.80 indicating that overall frailty is not a strong predictor of outcomes. These findings support the need for further research into preoperative risk assessment in pilon fracture patients.

III - retrospective cohort study.