Frailty measured by risk analysis index and adverse discharge outcomes after adult spine deformity surgery: analysis of 3104 patients from a prospective surgical registry (2011–2020)

Abstract

Background ContextMeasurement of frailty with the Risk Analysis Index (RAI) has demonstrated improved outcome prediction compared to other frailty indices across the surgical literature. However, the generalizability and clinical utility of preoperative RAI scoring for prediction of postoperative morbidity after adult spinal deformity surgery is presently unknown. Thus, recent studies have called for an RAI analysis of spine deformity outcomes. PurposeThe present study sought to evaluate the discriminatory accuracy of preoperative frailty, as measured by RAI, for predicting postoperative morbidity among adult spine deformity surgery patients using data queried from a large prospective surgical registry representing over 700 hospitals from 49 US states and 11 countries. Study Design/SettingSecondary analysis of a prospective surgical registry. Patient SampleAmerican College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database (2011–2020). Outcome MeasuresThe primary endpoint was “adverse discharge outcome” (ADO) defined as discharge to a non-home, non-rehabilitation nursing/chronic care facility. MethodsAdult spine deformity surgeries were queried from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database (2011–2020) using diagnosis and procedure codes. The relationship between increasing preoperative RAI frailty score and increasing rate of primary endpoint (ADO) was assessed with Cochran-Armitage linear trend tests. Discriminatory accuracy was tested by computation of concordance statistics (with 95% confidence interval [CI]) in receiver operating characteristic (ROC) curve analysis. ResultsA total of 3104 patients underwent spine deformity surgery and were stratified by RAI score: 0–10: 22%, 11–15: 11%, 16–20: 29%, 21–25: 26%, 26–30: 8.0%, 31–35: 2.4%, and 36+: 1.4%. The rate of ADO was 14% (N=439/3094). The rate of ADO increased significantly with increasing RAI score (p<.0001). RAI demonstrated robust discriminatory accuracy for prediction of ADO in ROC analysis (C-statistic: 0.71, 95% CI: 0.69–0.74, p<.001). In pairwise comparison of ROC curves (DeLong test), RAI demonstrates superior discriminatory accuracy compared to the 5-factor modified frailty index (mFI-5; p<.001). ConclusionPreoperative frailty, as measured by RAI, is a robust predictor of postoperative morbidity (measured by ADO) after adult spine deformity surgery. The frailty score may be translated directly to the bedside with a user-friendly risk calculator, deployed here: https://nsgyfrailtyoutcomeslab.shinyapps.io/spineDeformity