The Risk Analysis Index Has a Superior Predictive Threshold Than the Modified Frailty index-5 and Age for Predicting Poor Postoperative Outcomes in Neurosurgical Patients

Introduction: Recent quality improvement efforts have shifted away from overutilization of healthcare resources and toward enhancing overall patient care. Targeted initiatives highlight important patient factors that may influence length of hospital stay (LOS). Frailty, has been previously shown to significantly increase LOS. We sought to examine the predictive threshold of two frailty risk assessment instruments (risk assessment index [RAI], and modified frailty index-5 [mFI-5]), and age for extended LOS (eLOS), prolonged LOS (pLOS) and protracted LOS (prLOS).

Methods: All patients who underwent a neurosurgical procedure and entered into the American College of Surgeons NSQIP registry between 2012-2020 were included. Receiver operating characteristic (ROC) curve analyses were used to compare the area under the curve (AUC) RAI, mFI-5, and age for predicting our postoperative outcomes. eLos is defined as LOS ≥75 percentile, pLOS as ≥95 percentile, and prLOS as >30-day hospitalizations.

Results: A total of 411605 patients were included in this study. The median age was 59 years (IQR: 21) with a majority of female patients (52.2%), and white (75.2%). The RAI had a higher prediction threshold for 30-day eLOS AUC 0.653 than the mFI-5 AUC 0.552 and age AUC 0.573, P< 0.001. Similarly, pLOS (RAI AUC 0.718 than the mFI-5 AUC 0.568 and age AUC 0.559, P< 0.001) and prLOS (AUC 0.714 than the mFI-5 AUC 0.548 and age AUC 0.506, P< 0.001).

Conclusion : Longer hospital stays were predicted by higher levels of frailty, as evaluated by the RAI, which showed a superior predictive threshold to mFI-5 and age. Identification of frailty has the potential to improve perioperative risk assessment, risk mitigation, and the timely involvement of multidisciplinary teams to decrease length of stay.