Using the modified frailty index as a predictor of complications in adults undergoing transforaminal interbody lumbar...

To correlate the operative characteristics and complications of transforaminal lumbar interbody fusion (TLIF) to patient frailty status for the first time in a multicenter study.

Using the American College of Surgeons National Surgical Quality Improvement (ACS-NSQIP) database, all patients who underwent TLIF in 2015-2020 were filtered for their demographics, operative characteristics, and 30-day complication outcomes. Patients were stratified into 2 cohorts, low and high frailty, based on their modified frailty index 5 score. Univariate analysis was performed between the 2 cohorts for each collected variable, and multivariable analysis was performed to observe adjusted odds ratios (OR).

The frail cohort experienced more unplanned readmission (4.3 vs 6.6 %, p < 0.001). During hospital stays, the frail cohort experienced more overall complications (9.8 vs 13.8 %, p < 0.001). In contrast to the low frailty cohort, the high frailty patients saw longer hospital stays (3.27 vs. 3.69 days, p < 0.001). The high frailty group saw more discharges to an institution beside their home (89.6 vs 77.9 %, p < 0.001). Rates of superficial and deep surgical site infection, organ space infection, wound dehiscence, reintubation, renal insufficiency, urinary tract infection, stroke, cardiac arrest, DVT, sepsis, and septic shock were not significantly different. Multivariable analyses showed high frailty status as an independent predictor of unplanned readmissions, major complications, and preventing discharge to home.

mFI-5 serves as an effective predictor of surgical outcomes following TLIF and independently predicts unplanned readmission, discharge to home, and major complications. Noninfectious outcomes were more likely to be significantly different between the high- and low frailty groups, while all infectious outcomes apart from superficial surgical site infection and pneumonia were not significantly different between the cohorts.