Which lumbar spinal stenosis patients will improve with nonsurgical treatment? A secondary analysis of a randomized controlled trial

Abstract Background Lumbar spinal stenosis (LSS) can be disabling and is a leading reason for spinal surgery in older adults. While nonsurgical treatments are recommended as first-line treatment, it remains unclear which patients will benefit most. Purpose To identify patient characteristics associated with larger improvements or larger treatment effects among adults receiving nonsurgical LSS interventions. Design Secondary analysis of a randomized controlled trial. Setting Outpatient research clinics. Subjects 216 older adults with symptomatic LSS. Methods Participants, recruited from November 2013 to June 2016, were randomized to receive: (1) manual therapy with an individualized exercise program (MTE); (2) a group exercise (GE) program; or (3) medical care (MC). We evaluated the association of baseline characteristics with 2-month change in primary outcomes: symptoms and function on the Swiss Spinal Stenosis questionnaire (SSSQ); and walking capacity in meters (m) on the self-paced walking test (SPWT). Baseline characteristics included sociodemographic and clinical variables. To explore heterogeneity of treatment effects, we evaluated unadjusted stratified estimates when comparing MTE to GE/MC. Additionally, we included an interaction term in models to test for statistical interaction. Results At baseline, participants (mean age = 72, 54% female, 23% non-white) had moderate LSS-related symptoms/impairment (mean SSSQ score = 31.3) and limited walking capacity on SPWT (mean = 451 m). The overall improvement on SSSQ was 2.5 points with larger improvements observed among younger, non-white, non-smoking participants, and those with worse baseline LSS or back-related symptoms/impairment. Overall improvement on the SPWT was 205 m with larger improvements observed among younger participants, those with higher baseline physical activity levels and participants without knee osteoarthritis. For SSSQ, the treatment effect was larger among adults aged < 70 versus older adults (MTE vs. GE/MC; mean difference MD = − 4.06, 95% CI = − 6.29 to − 1.83 vs. MD = − 0.47. 95% CI = − 2.63 to 1.69, respectively; p -for-interaction = 0.02). For walking capacity, the treatment effect was larger among adults with hip osteoarthritis compared to those without (MTE vs. GE/MC; MD = 500 m, 95% CI = 71 to 929, vs MD = 13 m, 95% CI = − 120 to 147, respectively; p -for-interaction = 0.007). Conclusions In a sample receiving nonsurgical treatments for LSS, we identified patient-level characteristics associated with larger improvements and/or treatment effects. If confirmed in larger randomized controlled trials, these findings may guide clinical decision-making to enhance clinical outcomes. ClinicalTrials.gov identifier NCT01943435.