Phenotypes of patients with symptomatic lumbar spinal stenosis presenting for non-operative care: Baseline data from The Lumbar Stenosis Prognostic Subgroups for Personalizing Care and Treatment (PROSPECTS) Study.

Abstract

Objective: This study describes the enrollment and baseline characteristics of participants in the PROSPECTS cohort and explores subgroups of patients presenting for non-operative care.

Design: Cross-sectional study.

Setting and subjects: We enrolled adults ≥50 years initiating non-operative care for symptomatic lumbar spinal stenosis. We excluded those with serious spinal pathology, conditions limiting ambulation, and prior or planned lumbar surgery.

Methods: We collected demographics, the PROMIS-29, pain intensity, Oswestry Disability Index, Swiss Spinals Stenosis Questionnaire, chronicity of symptoms, pain sites, comorbidities, falls, and opioid use. We used descriptive statistics to characterize the sample and Latent Class Analysis (LCA) to derive subgroups with distinct phenotypes. The best model was selected based on model fit statistics, class separation, and clinical interpretability.

Results: We enrolled 598 participants. The mean age was 67 (SD = 9), and 61% were female. Back and leg pain were present for ≥1 year for 65% of participants. Multiple pain sites were common, mean of 4.3 sites (SD = 2.2), and a majority of patients had multiple comorbidities (54%). We selected a four-class solution as the best model from the LCA. These phenotypes were described as 1) "high pain impact, low psychosocial features" n = 233 (39%), 2) "mild pain impact, low psychosocial features" n = 218 (36%), 3) "high pain impact, complex health needs" n = 95 (16%), and 4) "acute, intermittent, moderate-severe leg pain with high pain impact" n = 52 (9%).

Conclusions: These phenotypes reflect distinct profiles which may inform health needs and patient-centered care. Future studies should examine longitudinal outcomes to establish their clinical utility and prognostic value.