Frailty as an Independent predictor of perioperative risk and recovery in long-segment thoracolumbar fusion surgery: A retrospective cohort study.

Background: Frailty, a multifactorial condition reflecting diminished physiological reserve, is associated with worse surgical outcomes and recovery, particularly in complex surgical procedures. This study aims to investigate frailty, as measured by the mFI-5, as a predictor of perioperative risk and recovery in patients undergoing long-segment thoracolumbar fusion for adult spinal deformity (ASD). Methods: We conducted a retrospective analysis to identify patients undergoing long-segment thoracolumbar fusion for ASD. Our primary outcome measures included anesthesia duration, estimated blood loss (EBL), intraoperative blood transfusion volumes, ICU stay duration, hospital length of stay (LOS), time to ambulation, postoperative complications, readmission rates, and discharge disposition. Patients were categorized as not frail (mFI-5 = 0), partially frail (mFI-5 = 1), or frail (mFI-5 ≥ 2). ANOVA and chi-square tests examined group differences, while multivariable regression, adjusting for age, gender, BMI, and number of levels fused, assessed frailty's impact on perioperative outcomes. Results: About 235 patients were included; 45.1% were frail, 34.5% were partially frail, and 20.4% were not frail. In multivariable regression analysis, each unit increase in frailty score was associated with longer anesthesia duration (β = 11.1 minutes, p = .023), greater EBL (β = 150.5 mL, p = .026), and higher intraoperative blood transfusion volumes (β = 228.2 mL, p = .002). Frailty also independently predicted prolonged ICU stays (β = 9.0 hours, p = .002), increased hospital LOS (β = 0.6 days, p = .015), and delayed time to ambulation (β = 0.4 days, p = .039). Postoperative complications, readmission rates, and discharge disposition were not significantly different among frailty groups. Conclusions: Frailty independently predicts longer anesthesia duration, greater EBL, increased intraoperative transfusion requirements, and significantly delayed postoperative recovery in ASD patients, including prolonged ICU stays, extended hospital LOS, and delayed ambulation. While targeted perioperative strategies-such as optimizing preoperative comorbidities and enhancing intraoperative monitoring and rehabilitation-may not reduce a patient's underlying frailty status, our findings suggest that early identification of frailty can help tailor perioperative planning, anticipate recovery trajectories, and support informed decision-making for clinicians and patients.