Associations between surgeons' preoperative expectations of lumbar surgery and patient-reported 2-year outcomes.

Abstract

Surgeons' preoperative expectations of lumbar surgery may be associated with patient-reported postoperative outcomes. Preoperatively spine surgeons completed a validated Expectations Survey for each patient estimating amount of improvement expected (range 0-100). Preoperative variables were clinical characteristics, spine-specific disability (ODI), and general health (RAND-12). Two years postoperatively patients again completed these measures and global assessments of satisfaction. Surgeons' expectations were compared to preoperative variables and to clinically important pre- to postoperative changes (MCID) in ODI, RAND-12, and pain and to satisfaction using hierarchical models. Mean expectations survey score for 402 patients was a 57 (IQR 44-68) reflecting moderate expectations. Lower scores were associated with preoperative older age, abnormal gait, sensation loss, vacuum phenomena, foraminal stenosis, prior surgery, and current surgery to more vertebrae (all p ≤ .05). Lower scores were associated postoperatively with not attaining MCID for the ODI (p = .02), RAND-12 (p = .01), and leg pain (p = .01). There were no associations between surgeons' scores and satisfaction (p = .06-.27). 55 patients (14%) reported unfavorable global outcomes and were more likely to have had fracture/infection/repeat surgery (OR 3.2, CI 1.6-6.7, p = .002). Surgeons' preoperative expectations were associated with patient-reported postoperative improvement in symptoms and function, but not with satisfaction. These findings are consistent with clinical practice in that surgeons expect some but not complete improvement from surgery and do not anticipate that any particular patient will have markedly unfavorable satisfaction ratings. In addition to preoperative discussions about expectations, patients and surgeons should acknowledge different types of outcomes and address them jointly in postoperative discussions.