The risk analysis index demonstrates exceptional discrimination in predicting frailty's impact on neurosurgical length of stay quality metrics

• Published in: World journal of surgery Vol. 48; no. 1; pp. 59 - 71
• Main Authors: Gagliardi, Thomas A., Conti, Joseph T., Courville, Jordyn T., Owodunni, Oluwafemi P., Courville, Evan N., Kazim, Syed F., Schmidt, Meic H., Bowers, Christian A.
• Format: Journal Article
• Language: English
• Published: United States 01-01-2024
• Subjects: Adult; Age Factors; Aged; Female; frailty; Frailty - diagnosis; hospital length of stay; Humans; Length of Stay - statistics & numerical data; Male; Middle Aged; modified frailty index‐5; Neurosurgical Procedures - statistics & numerical data; Quality Indicators, Health Care; quality metrics; Retrospective Studies; risk analysis index; Risk Assessment; frailty; hospital length of stay; risk analysis index; quality metrics; modified frailty index‐5
• ISSN: 0364-2313; 1432-2323
• DOI: 10.1002/wjs.12020

Background
Quality measures determine reimbursement rates and penalties in value‐based payment models. Frailty impacts these quality metrics across surgical specialties. We compared the discriminatory thresholds for the risk analysis index (RAI), modified frailty index‐5 (mFI‐5) and increasing patient age for the outcomes of extended length of stay (LOS [eLOS]), prolonged LOS within 30 days (pLOS), and protracted LOS (LOS > 30).
Methods
Patients ≥18 years old who underwent neurosurgical procedures between 2012 and 2020 were queried from the ACS‐NSQIP. We performed receiver operating characteristic analysis, and multivariable analyses to examine discriminatory thresholds and identify independent associations.
Results
There were 411,605 patients included, with a median age of 59 years (IQR, 48–69), 52.2% male patients, and a white majority 75.2%. For eLOS: RAI C‐statistic 0.653 (95% CI: 0.652–0.655), versus mFI‐5 C‐statistic 0.552 (95% CI: 0.550–0.554) and increasing patient age C‐statistic 0.573 (95% CI: 0.571–0.575). Similar trends were observed for pLOS‐ RAI: 0.718, mFI‐5: 0.568, increasing patient age: 0.559, and for LOS>30‐ RAI: 0.714, mFI‐5: 0.548, and increasing patient age: 0.506. Patients with major complications had eLOS 10.1%, pLOS 26.5%, and LOS >30 45.5%. RAI showed a larger effect for all three outcomes, and major complications in multivariable analyses.
Conclusion
Increasing frailty was associated with three key quality metrics that is, eLOS, pLOS, LOS > 30 after neurosurgical procedures. The RAI demonstrated a higher discriminating threshold compared to both mFI‐5 and increasing patient age. Preoperative frailty screening may improve quality metrics through risk mitigation strategies and better preoperative communication with patients and their families.