Digital pain diagrams to identify common lumbar spinal stenosis pain distribution patterns: a cross-sectional latent class analysis.

Abstract

To identify common pain distribution patterns on digital pain diagrams (PDs) in patients with lumbar spinal stenosis (LSS) presenting to a secondary spine care setting. Patients presenting to a secondary spine center with a diagnostic code for LSS and a completed PD were included. Latent class analysis of PDs was used to identify unique pain distribution patterns. The final class solution was selected based on the most suitable combination of the Bayesian Information Criterion, posterior probabilities, minimum class size, and the clinical interpretability of the LSS pain pattern distributions. A total of 2,379 patients with LSS were included. A six-class model of LSS pain distribution patterns was the preferred model solution. Class 1 (11.4%) represented a bilateral posterior leg pain (11.4%) distribution pattern; Class 2 (8.7%) bilateral posterior and anterior leg pain; Class 3 (26.1%) unilateral posterior leg pain; Class 4 (21.0%) unilateral posterior leg pain with low back pain; Class 5 (22.9%) unilateral anterior and posterior leg pain; and Class 6 (9.9%) multisite pain. We identified clinically recognizable pain distribution patterns in patients with LSS presenting to a secondary spine center in Denmark. These preliminary pain distribution patterns may represent clinical LSS phenotypes, but at minimum suggest that heterogenous pain presentations are common in people with LSS. Future work is needed to validate these pain distribution patterns in other LSS samples and determine their relationship with diagnostic and treatment outcomes.