Answers are grounded in this paper and cite their sources.
Study Design. Single-center retrospective cohort study. Objective. To identify independent predictors of SAVES-V2–graded complications after elective spine surgery, and whether frailty (mFI-5) and JOA score remain associated after adjusting for surgical region. Methods. A total of 691 adult patients undergoing elective spine surgery (cervical, thoracic, lumbar, or deformity; 2014–2024) were analyzed. The primary outcome was SAVES-V2 ≥3 complications (90 patients, 13.0%). Three multivariable models addressed hypertension/mFI-5 overlap (r=0.753): Model A (n687), Model B (hypertension excluded; n=687), and Model C (Model B plus region; n=635), designated primary as region improved fit ( P =0.036). Surgical levels’ functional form was assessed with a spline and clinically grouped categories. Results. Transient neurological deficit (n=41; C5 palsy, n=29) and surgical site infection requiring reoperation (n=19) were most frequent. A linear per-level specification was not well supported ( P <0.001); risk rose through 4–6 levels and plateaued thereafter (OR 13.89 for 4–6 vs. 1 level; P <0.001). Surgical levels were independently associated with complications in Model C (OR 1.281/level) and Model B (OR 1.258/level; both P <0.001). After region adjustment, JOA score, age, male sex, and BMI were no longer significant (all P >0.1); lumbar carried lower risk than cervical (OR 0.408; P =0.014). The mFI-5 remained paradoxically protective (OR 0.720; P =0.034). Discrimination was poor for both mFI-5 (C=0.487) and JOA score (C=0.582; DeLong P =0.007), a secondary finding; Model C showed the best discrimination overall (C=0.731). Conclusions. Number of surgical levels, not frailty or JOA score, remained the most robust predictor after region adjustment, though non-linear and not a comprehensive complexity measure. Both frailty and JOA showed limited discrimination, insufficient alone for perioperative risk stratification. Level of Evidence.
No takes yet. Share an insight, caveat, or question.
Kumagai et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: