Frailty (mFI-5) and Age Predict Medical Complications After Posterior Lumbar Interbody Fusion in Older Adults: A Retrospective Cohort Study

Objective: To evaluate the prognostic utility of the five-item modified Frailty Index (mFI-5) for postoperative outcomes in older adults undergoing posterior lumbar interbody fusion (PLIF) for degenerative lumbar disease and to develop an interpretable preoperative risk model for medical complications (CxME). Methods: We retrospectively reviewed consecutive patients aged ≥65 years who underwent PLIF for lumbar spondylosis at a single tertiary institution. Baseline demographics, comorbidities, symptoms, American Society of Anesthesiologists (ASA) physical status, bone mineral density, antithrombotic use, perioperative laboratory findings, and operative variables were collected. CxME were defined as Clavien–Dindo grade ≥ II complications occurring during index hospitalization or within 30 days postoperatively. mFI-5 was calculated from five preoperative variables and stratified as 0, 1, and ≥2. Multivariable logistic regression was used to identify independent predictors of CxME. Results: Among 255 patients (mean age 72.6 years), 53 (20.8%) developed CxME. Patients with CxME were older and had higher rates of diabetes mellitus and preoperative dependency. mFI-5 ≥ 2 was more frequent in patients with CxME than in those without (66.0% vs. 27.3%, p < 0.001). Higher frailty was associated with older age, greater comorbidity burden, higher ASA class, lower preoperative hemoglobin, greater transfusion exposure, longer hospital stay, and a higher incidence of CxME (14.0%, 9.6%, and 38.9% for mFI-5 scores 0, 1, and ≥2, respectively; p < 0.001). In multivariable analysis, age (OR 1.133, 95% CI 1.055–1.216; p < 0.001) and mFI-5 (OR 2.103, 95% CI 1.387–3.188; p < 0.001) independently predicted CxME. The Age + mFI-5 model showed fair discrimination (optimism-corrected area under ROC 0.734). Conclusions: Preoperative frailty and age independently predicted postoperative medical complications after PLIF in older adults. The Age + mFI-5 model may support risk stratification, counselling, and perioperative optimization.