Preoperative frailty risk in deep brain stimulation patients: Risk analysis index predicts Clavien-Dindo IV complications

Abstract

ObjectiveDeep brain stimulation (DBS) improves patients’ quality of life in multiple movement disorders and chronic neurodegenerative diseases. There are no published studies assessing frailty’s impact on DBS outcomes. We evaluated frailty’s impacts on DBS outcomes, comparing discriminative thresholds of the risk analysis index (RAI) to modified frailty index-5 (mFI-5) for predicting Clavien-Dindo complications (CDIV). MethodsPatients who underwent DBS between 2015 and 2019 in the ACS-NSQIP registry were included. We employed receiver operating characteristic (ROC) curve to examine the discriminative thresholds of RAI and mFI-5 and multivariable analyses for postoperative outcomes. Our primary outcome was CDIV, and secondary outcomes were discharge to higher-level care facility, unplanned reoperation within 30 days, in any hospital, for any procedure related to the index procedure, and extended length of stay. ResultsA total of 3795 patients were included. In the ROC analysis for CDIV, RAI showed superior discriminative threshold (C-statistic = 0.70, 95% CI 0.61–0.80, <0.001) than mFI-5 (C-statistic = 0.60, 95% CI 0.49–0.70, P = 0.08). On multivariable analyses, frailty stratified by RAI, had independent associations with CDIV, i.e., pre-frail 2-fold increase OR 2.04 (95% CI: 1.94–2.14) p < 0.001, and frail 39% increase OR 1.39 (95% CI: 1.27–1.53), p < 0.001. ConclusionFrailty was an independent risk-factor for CDIV. The RAI had superior discriminative thresholds than mFI-5 in predicting CDIV after DBS. Our ability to identify frail patients prior to DBS presents a novel clinical opportunity for quality improvement strategies to target this specific patient population. RAI may be a useful primary frailty screening modality for potential DBS candidates.