New Surgical Frailty Scoring Tool: Modified 4-Factor Functional Frailty Index

Abstract

Background: Frailty is the age-related decline in functional reserve, resulting in physical and disease vulnerability. The established 5-factor modified frailty index (mFI-5), based on the NSQIP database, is predictive but does not encompass core components of frailty-namely physical abilities, cognitive status, and nutrition. Therefore, we sought to establish a scoring index more indicative of the frailty phenotype.

Study design: Our retrospective study included patients 75 years of age or older, from NSQIP 2021 to 2022. The modified 4-factor functional frailty index (mFF-4) is calculated by summing each present variable: previous falls, dementia, low BMI, and nonindependent functional status. Patients were stratified into frailty levels: none (mFF-4 = 0), intermediate (mFF-4 = 1), and high (mFF-4 = 2+). Area under the curve (AUC) and Akaike information criterion values assessed predictability relative to the reference mFI-5. Odds ratios (ORs) for 30-day outcomes were calculated using logistic regression.

Results: The sample included 265,041 patients. The median age was 79 years (interquartile range 77 to 84 years). The mFF-4 AUCs were consistently higher than mFI-5; mFF-4 AUCs were 0.8679 for mortality, 0.7298 for postoperative complications, and 0.7645 for geriatric outcomes. Delta (Akaike information criteria) values exceeded 2, favoring mFF-4 models. Compared with nonfrail, mFF-4 = 2+ had increased odds of mortality (OR 2.47, 95% CI 2.31 to 2.63), postoperative complications (OR 1.37, 95% CI 1.32 to 1.43), and geriatric outcomes (OR 4.54, 95% CI 4.30 to 4.79).

Conclusions: The new mFF-4 scoring tool embodies the true frailty phenotype to accurately predict 30-day outcomes across surgical subspecialties and the overall population. It can serve as a new NSQIP "frailty" definition while also providing a framework for surgeons and patients alike to conceptualize the elevated risks it poses.