Background Hip fractures present a significant public health challenge, with between 260,000 to 300,000 admissions for hip fractures in the US and projections of more than 500,000 per year by 2040. The rising incidence and substantial burden on healthcare systems necessitate proper evaluation and management of these already frail patients as hip fractures are associated with increased mortality and morbidity. Frailty scales have been used extensively within the orthopedic literature as a potential preoperative risk stratification tool. This study aims to compare the effectiveness of frailty as measured by the Revised Risk Analysis Index (RAI-rev) and the 5-Factor Modified Frailty Index (mFI-5) in predicting 30-day postoperative outcomes in surgically managed hip fracture.

Methods The ACS-NSQIP database was queried for patients with a diagnosis of traumatic hip fracture undergoing surgical intervention from (2011-2020). Frailty was assessed using the RAI-rev and mFI-5. Thirty-day outcomes included mortality, extended length of stay (eLOS) and non-home discharge (NHD). Multivariate analysis was conducted to assess predictive value of frailty scales, presented as odds ratio (OR). Discriminatory accuracy was assessed using Receiver operating characteristic (ROC) curve and quantified using C-statistic. Significance was determined by p<0.05.

Results There were 160,446 patients included with a median age of 83 years, with 68.9% females and 91.86% white. The RAI-rev classified 16,175 patients as robust, 89,815 as normal, 46,531 as frail, and 7,925 as very frail. Multivariate analysis revealed increasing frailty as measured by the RAI-rev had greater odds for 30-day mortality (normal: OR 2.10 (1.73 -2.59), frail: OR 5.35 (4.35 - 6.63), very frail: OR 11.32 (9.13 - 14.13), p<0.001) than the mFI-5 (prefrail: OR 1.39 (1.29 - 1.50), frail: OR 2.15 (1.99 - 2.32) severely frail: OR 3.70 (3.39 -4.04) , p<0.001). ROC analysis demonstrated that compared to the mFI-5, the RAI-rev had superior discriminatory accuracy for 30-day mortality (RAI: 0.74, mFI-5: 0.62, p<0.001), eLOS (RAI-rev: 0.57, mFI-5: 0.56, p<0.001), and NHD (RAI-rev: 0.57, mFI-5: 0.54, p<0.001).

Conclusion The RAI-rev demonstrated superior predictive capability and discriminatory accuracy for assessing 30-day mortality, eLOS, and NHD, when compared to the mFI-5. These findings support the use of the RAI-rev in preoperative risk assessment for hip fractures in an increasingly frail population. In doing so, surgeons may identify high risk patients missed without obvious multimorbidity, thus reducing postoperative morbidity and mortality in patients with traumatic hip fracture.