Primary care involvement and outcomes of care in patients with a workers' compensation claim for back pain

Abstract

Study design: A retrospective case series of primary care patients receiving Workers' Compensation (WC) for back pain.

Objective: To describe the extent of primary care provider (PCP) involvement in work-related low back pain (WRLBP), and whether PCP involvement is a significant determinant of patterns of care or outcomes.

Summary of background data: In general, WRLBP has inferior outcomes compared with nonoccupational LBP. Although it has been suggested that better outcomes are achieved when care for WRLBP involves specialists such as occupational medicine physicians rather than a PCP, limited evidence supports this claim.

Methods: Adult patients were identified by electronic records from four hospital-affiliated PCP practices in fiscal years 1996 to 1998. Those with a WRLBP claim filed during the same time period were identified by searching a WC claims database. Medical records were reviewed to assess past medical history, the extent of prior PCP contacts for any visits before and for visits 2 years after the reported onset of the WRLBP claim, and detailed information on visits for this condition. Disability outcomes (total costs and days of work disability) were obtained from the WC database.

Results: Among 68,710 individuals with a PCP practice visit, 118 with a WRLBP claim were identified. The final study sample included 87 patients with a WRLBP claim and at least 1 documented PCP practice visit during the study period. Seventy-eight percent of patients saw the PCP at least once in the 2 years after the claim onset date, but only 34 patients (39%) visited the PCP for their WRLBP, and 20 (23%) had more than 1 visit. Although almost all patients with a PCP visit for WRLBP saw a physician (not the PCP in 79% of cases) within a week of the claim onset date, the median number of days between the onset date and the first WRLBP visit to the PCP was 47.5. Patients with a prior history of back problems were more likely to have a PCP visit for WRLBP (odds ratio 2.9, 95% confidence interval 1.1-7.7). Patients with PCP visits for WRLBP had higher total and medical costs than those without PCP visits, but a similar number of paid disability days. After controlling for other potential predictors, involvement of the PCP was not a significant predictor of the total cost of the WRLBP claim.

Conclusions: Many individuals with a WC claim do not have a stable PCP relationship around the onset of their WRLBP episode. Those with PCP relationships uncommonly involve the PCP in their WRLBP, and if they do, it is usually later in the course because of persistent or recurrent symptoms. Disability outcomes appear to be similar regardless of whether a PCP was involved. However, PCP involvement is associated with greater medical costs, which may reflect confounding based on duration of symptoms rather than the nature of the care provided. Additional research is needed to understand how different patterns of care for patients with a WRLBP claim relate to outcomes and how these patients compare with individuals without such a claim, as well as the factors that lead patients to involve their PCP or not.