The Weight of Frailty in Neurosurgery Patients: Analyzing the Combined Effect of Frailty and Body Mass Index on 30-Day Postoperative Mortality

Abstract

BackgroundThere is a rising prevalence of overweight and obese persons in the US, and there is a paucity of information about the relationship between frailty and body mass index (BMI). Therefore, we examined discrimination thresholds and independent relationships of the risk analysis index (RAI), modified frailty index-5 (mFI-5) and increasing patient age in predicting 30-day postoperative mortality. MethodsThis retrospective American College of Surgeons National Surgical Quality Improvement Program analysis compared all overweight or obese adult patients who underwent neurosurgery procedures between 2012-2020. We compared discrimination using receiver operating characteristic (ROC) curve analysis for RAI, mFI-5, and increasing patient age. Furthermore, multivariable analyses, as well as subgroup analyses by procedure type i.e., spine, skull base and other (vascular and functional) were performed, and reported as odds ratios (OR) and 95% confidence intervals (CI). ResultsWe included 315,725/412,909 (76.5%) neurosurgery patients, with a median age of 59 years (IQR:48-68), predominately white 76.7% and male 54.3%. ROC analysis for 30-day postoperative mortality demonstrated a higher discriminatory threshold for RAI (C-statistic:0.790, 95%CI:0.782-0.800) compared to mFI-5 (C-statistic:0.692, 95%CI:0.620-0.638) and increasing patient age (C-statistic:0.659, 95%CI:0.650-0.668). Multivariable analyses showed a dose-dependent association and a larger magnitude of effect by RAI: frail patients OR:11.82 (95%CI:10.57-13.24), and very frail patients OR:31.19 (95%CI:24.87-39.12). A similar trend was observed in all subgroup analyses i.e., spine, skull base, and other (vascular and functional) procedures (p≤0.001). ConclusionIncreasing frailty was associated with a higher rate of 30-day postoperative mortality, with a dose-dependent effect. Furthermore, the RAI had a higher threshold for discrimination and larger effect sizes than mFI-5 and increasing patient age. These findings support RAI’s use in preoperative assessments, as it has the potential to improve postoperative outcomes through targeted interventions.