135 Biological Age Outperforms Frailty Scores for Predicting 30-Day Outcomes After Neurosurgery

View at publisher Summarise with VIRAT
Authors: Ritvik Jillala, Arjun Kumar Menta, Carlos Andres Aude, Tej Deepak Azad, Patrick J. Kramer, Vikas N. Vattipally, Antony A. Fuleihan, Daniel Davidar, Meghana Bhimreddy, Joseph M. Dardick, Nicholas Theodore
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

INTRODUCTION: Chronological age and checklist frailty indices incompletely capture the physiological reserve that governs neurosurgical risk. PhenoAge, a laboratory-based biological-age metric, shows promise in surgical cohorts but remains untested in neurosurgical patients. METHODS: Adult cranial and spinal operations in the ACS-NSQIP (2016-2020) were analyzed. PhenoAge was calculated from nine pre-operative laboratory values, trained from data within the NHANES dataset; frailty scores followed published algorithms. Endpoints were 30-day major morbidity and all-cause mortality. Multivariable logistic and Cox models were adjusted for demographics, ASA class, urgency, and procedure category. Models were compared using C-indices, likelihood-ratio Δχ2, categorical net reclassification improvement (NRI), and Boruta random-forest variable importance. RESULTS: The cohort comprised 127,375 patients (median age 61y, 49% female); 13% experienced major morbidity and 1.2% died within 30 days. PhenoAge outperformed chronological age for 30-day outcomes—mortality: HR 1.91, C-index 0.906, LR χ2 = 157 vs age HR 1.71, C-index 0.902; morbidity: HR 1.19, C-index 0.720, LR χ2 = 250 vs age HR 1.10, C-index 0.717—and, when both entered, age lost prognostic value (mortality HR 0.79, morbidity HR 0.67). The highest PhenoAge decile showed 3.1× more complications and 25.2× higher mortality than the lowest. PhenoAge dominated mortality prediction (C-index: 0.905, Δχ2: 1469) vs RAI 0.901/303, rev-RAI 0.901/295, mFI-5 0.900/54, achieving NRI +52%—at least 40% higher than any comparator; morbidity results paralleled this (0.720/4253 vs 0.716/568, 0.716/628, 0.716/498; NRI +26%). PhenoAge ranked second only to ventilator dependence in variable-importance analysis. CONCLUSIONS: A readily-automated biological-age metric outperforms traditional frailty instruments in short-term neurosurgical outcome prediction and meaningfully reclassifies patient risk. incorporating PhenoAge into pre-operative workflows could refine surgical decision-making and resource allocation.

Fields of Study & Subjects

Frailty in Older Adults Cardiac Anesthesia and Surgical Outcomes Total Knee Arthroplasty Outcomes

Publication details

Journal / source
Neurosurgery
Volume
72
Issue
Supplement_1
Pages
21–21
Publication date
26 March 2026
Language
EN