Prospective validation and application of the Trauma-Specific Frailty Index: Results of an American Association for the Surgery of Trauma multi-institutional observational trial
Bellal Joseph,Abdul Tawab Saljuqi,Joseph D Amos,Amanda Teichman,Melissa L Whitmill,Tanya Anand,Hamidreza Hosseinpour,Sigrid K Burruss,Julie A Dunn,Kaveh Najafi,Laura N Godat,Toby M Enniss,Thomas H Shoultz,Tanya Egodage,Tasce Bongiovanni,Joshua P Hazelton,Kristin P Colling,Todd W Costantini,Deborah M Stein,Thomas J Schroeppel,Jeffry Nahmias,AAST Frailty MIT Study Group,Khaled El-Qawaqzeh,Rachel L Choron,Paul B Comish,Kyle Leneweaver,Brandi Palmer,Michael S Truitt,Mike Farrell,Lacee J Laufenberg,Erica Lasso-Tay,Zachery Stillman,Daniel T Hass,Heather M Grossman,Darnell Gordon,Cassandra Krause,Jonathan Thomas
DOI: https://doi.org/10.1097/TA.0000000000003817
2023-01-01
Abstract:Background: The frailty index is a known predictor of adverse outcomes in geriatric patients. Trauma-Specific Frailty Index (TSFI) was created and validated at a single center to accurately identify frailty and reliably predict worse outcomes among geriatric trauma patients. This study aims to prospectively validate the TSFI in a multi-institutional cohort of geriatric trauma patients. Methods: This is a prospective, observational, multi-institutional trial across 17 American College of Surgeons Levels I, II, and III trauma centers. All geriatric trauma patients (65 years and older) presenting during a 3-year period were included. Frailty status was measured within 24 hours of admission using the TSFI (15 variables), and patients were stratified into nonfrail (TSFI, ≤0.12), prefrail (TSFI, 0.13-0.25), and frail (TSFI, >0.25) groups. Outcome measures included index admission mortality, discharge to rehabilitation centers or skilled nursing facilities (rehab/SNFs), and 3-month postdischarge readmissions, fall recurrences, complications, and mortality among survivors of index admission. Results: A total of 1,321 geriatric trauma patients were identified and enrolled for validation of TSFI (nonfrail, 435 [33%]; prefrail, 392 [30%]; frail, 494 [37%]). The mean ± SD age was 77 ± 8 years; the median (interquartile range) Injury Severity Score was 9 (5-13). Overall, 179 patients (14%) had a major complication, 554 (42%) were discharged to rehab/SNFs, and 63 (5%) died during the index admission. Compared with nonfrail patients, frail patients had significantly higher odds of mortality (adjusted odds ratio [aOR], 1.93; p = 0.018), major complications (aOR, 3.55; p < 0.001), and discharge to rehab/SNFs (aOR, 1.98; p < 0.001). In addition, frailty was significantly associated with higher adjusted odds of mortality, major complications, readmissions, and fall recurrence at 3 months postdischarge ( p < 0.05). Conclusion: External applicability of the TSFI (15 variables) was evident at a multicenter cohort of 17 American College of Surgeons trauma centers in geriatric trauma patients. The TSFI emerged as an independent predictor of worse outcomes, both in the short-term and 3-month postdischarge. Level of evidence: Prognostic and Epidemiological; Level III.