Does Disordered Sleep Moderate the Relationship Between Pain, Disability and Downstream Health Care Utilization in Patients With Low Back Pain? A Longitudinal Cohort From the US Military Health System

Study Design. Prospective cohort. Objective. The purpose of this study was to evaluate the influence of disordered sleep on the relationship between pain and health care utilization (HCU) and pain-related disability and HCU in individuals with low back pain (LBP). Summary of Background Data. Disordered sleep and pain influence LBP outcomes, but their relationship with health care seeking after an episode of LBP has not been investigated and could help identify who is at risk for long-term medical care. Methods. This study included patients with LBP participating in a self-management class at a large US military hospital between March 1, 2010 and December 4, 2012. Pain intensity, disability (Oswestry Disability Index), and sleepiness (Epworth Sleepiness Scale) were captured at baseline. Medical visits for a sleep disorder in the 12 months before the class and LBP-related healthcare utilization for the 12 months following the class were abstracted from the Military Health System Data Repository. Separate multivariate analyses evaluating pain intensity and disability as predictors of HCU were developed, with sleepiness and the presence of a sleep disorder as potential moderators. Analyses were adjusted for age, sex, history of back pain, and mental health comorbidities. Results. A total of 757 consecutive participants were included, with 195 (26.8%) diagnosed with a subsequent sleep disorder. Sleepiness was not a significant predictor of HCU. The main effects of disability, pain intensity, and presence of a sleep disorder were significant across all analyses, with higher disability, pain intensity, and presence of a sleep disorder associated with higher predicted visits and costs for LBP. The presence of a sleep disorder was not a significant moderator in any model. Conclusion. Higher pain intensity and disability predicted higher pain-related HCU in the year following a LBP self-management class. The presence of a sleep disorder diagnosis, as recorded in medical records, had a significant independent effect on LBP-related health care visits and costs beyond the influences of pain intensity, disability, and other key demographic and health-related characteristics, but did not moderate these relationships.