Back pain among nursing home workers: Persistence within individuals

Background and Objective: Healthcare workers have high rates of back pain. Within a large chain of nursing homes, we found reductions in biomechanical load and cost of compensation claims following implementation of a safe resident handling program (SRHP). We have now examined whether back pain frequency also changed over time. Methods: Surveys were conducted pre-SRHP (baseline) and 3, 12, and 24 months after. The outcome was low back pain (LBP - dichotomous) in the previous three months. Multivariable Glimmix modeling included time as a fixed effect, adjusted for individual subject as a random effect, and examined clustering by region or facility as a fixed effect. Results: A total of 692 individuals completed the baseline plus at least one other survey. Among those with baseline LBP (n=215), 64% had pain in their next follow-up survey. In those without baseline LBP (n=459), 17% had pain on next follow-up. The crude odds ratio was 8.5 (95% CI 5.9-12.4) for LBP at follow-up, conditional on baseline LBP. Back pain prevalence was fairly stable over the two-year study period. Glimmix models (Poisson response) gave coefficients close to 1.0 for calendar time (months from baseline). In a logit model (binomial response), the odds ratio for follow-up LBP given baseline LBP was 10.5 (p<0.0001). Discussion/Conclusions: LBP was highly persistent over time in this fixed cohort. Despite reduced workers' compensation claim rates for back disorders, symptom prevalence did not markedly decrease. Future analyses, e.g., of compensation claim-filing, will attempt to resolve the apparently discrepant findings from this study.