A predictive tool incorporating frailty scores for perioperative risk assessment in patients with spinal metastasis: a national retrospective cohort study.
TL;DR
Integrating serum albumin, hematocrit, BMI, spinal region, and procedure type with frailty indices enhances risk stratification and informs surgical planning, preoperative optimization, and patient counseling in this vulnerable population of patients.
Abstract
Background
Surgical treatment for spinal metastases is associated with high perioperative risk due to tumor burden, neurologic compromise, and limited physiological reserve. Frailty is a recognized predictor of adverse outcomes, yet most existing indices exclude key clinical variables. This study evaluated and enhanced the predictive performance of three frailty instruments-the 5-item Modified Frailty Index (mFI-5), 11-item mFI (mFI-11), and Risk Analysis Index-Administrative (RAI-A)-by incorporating five clinically relevant covariates: serum albumin, hematocrit, body mass index (BMI), spinal region, and procedure type.
Methods
A retrospective cohort analysis was performed using the ACS-NSQIP database (2010-2022) to identify adult patients who underwent surgery for spinal metastases. Frailty scores and clinical covariates were analyzed using logistic regression and ensemble machine learning models. Primary outcomes were 30-day postoperative complications, readmission, and reoperation. Model performance was assessed using discrimination and reclassification metrics.
Results
A total of 5,052 patients were included. Higher frailty scores were significantly associated with increased complication rates, prolonged hospitalization, higher early mortality, and lower rates of discharge to home (p < 0.001). In multivariable analysis, vertebrectomy/corpectomy, thoracic or multilevel surgery, and lower preoperative hematocrit were independently associated with increased complication risk. Higher serum albumin was protective against early reoperation, while steroid use and delayed time to surgery predicted higher readmission risk. The RAI-A model performed best for complications, while the mFI-5 model showed strongest association with reoperation. Augmenting frailty indices with the five clinical covariates improved predictive performance across all outcomes.
Conclusion
Frailty is a clinically meaningful predictor of perioperative risk in spinal metastasis surgery. Integrating serum albumin, hematocrit, BMI, spinal region, and procedure type with frailty indices enhances risk stratification and informs surgical planning, preoperative optimization, and patient counseling in this vulnerable population. Our web-based risk calculator can aid in risk prediction while maximizing utility and simplicity: https://huggingface.co/spaces/Lansaol/Frailty_in_Spine_Met .