Patient Variability Drives Postoperative Outcome Volatility More Than Surgeon or Indication: A Bayesian Simulation Study of PROMIS Global Health for Lumbar Spinal Stenosis.

Background and objectives: Patient-reported outcome measures (PROMs) are ubiquitously used to assess surgical success after surgery for lumbar spinal stenosis (LSS); however, variability of PROMs can limit their utility as stand-alone outcomes. This study aimed to quantify the extent of PROM volatility after surgery for LSS and the relative contributions of patient and surgeon heterogeneity to that volatility.

Methods: Patients undergoing surgery for LSS between 2015 and 2023 with completed Patient-Reported Outcome Measurement Information System Global Health physical summary scores (PROMIS-PH) were retrospectively queried from a large prospectively collected institutional database. PROM volatility was measured by reversal fluctuations greater than the minimally clinically important difference in PROMIS-PH (5+ T-score points), deemed "reversal of surgical effects" (RSEs), over time with respect to patient's preoperative baseline (B-RSEs), and sequential postoperative values (S-RSEs). Bayesian simulation modeling quantified the relative contribution of patient (i.e., intrinsic factors such as demographics and psychological distress) and surgeon (i.e., extrinsic factors such as surgeon experience) variability on outcome volatility.

Results: Of 2439 included patients, 40% and 58% of patients experienced at least 1 B-RSEs and S-RSE, respectively, in the 2-year postoperative period, with peak incidence between 6 and 12 months postoperatively. Increased patient-level variability was associated with better peak postoperative PROMIS-PH. Patient-level outcome variability was significantly more responsible than surgeon-level variability for total outcome volatility (SD = 4.20 [95%CI: 4.03, 4.38] vs SD = 2.44 [95%CI: 1.78-3.20]). Eliminating variability attributed to surgeon did not significantly change postoperative outcome volatility for PROMIS-PH. Reducing patient variability did significantly reduce postoperative outcome volatility for PROMIS-PH.

Conclusion: PROMs are most volatile during the first postoperative year after surgery for LSS. Patient-level variability is predictive of better peak PROM scores postoperatively. Reducing patient, but not surgeon variability, reduced overall postoperative outcome volatility.