Effects of provider networks on health care costs for workers with short-term injuries

OBJECTIVES. This study examines the effects of preferred provider networks on health care costs and service utilization in the treatment of work-related injuries. RESEARCH DESIGN. A retrospective comparison of workers' compensation claims treated by network and nonnetwork providers was conducted. Pairwise matches of individual cases are used to control for differences in case mix and severity of injury between network patients and a nonnetwork comparison group. Cost differentials are separated into a price effect, the difference in costs attributed to network price discounts, holding services constant; and a utilization effect, the difference in costs attributed to differences in service utilization, holding prices constant. SUBJECTS. Data include approximately 87,000 workers' compensation claims, from California, Connecticut, and Texas, with injury dates between 1995 and 1997. The samples are restricted to five common injury types and work absences of less than 7 days. Workers treated solely by network providers are compared with a matched group of workers treated solely by nonnetwork providers. RESULTS. Average health care costs are lower for network claims than for matched nonnetwork claims. Price discounts explain a large part of the cost differentials for all injury groups studied, but differences in service utilization are also important for back cases and cumulative stress injuries. CONCLUSIONS. Networks bring the traditionally high costs of health care for work-related injuries closer to the costs of health care for off-the-job injuries. The network savings primarily reflect price discounts for the same services, thereby representing an increase in the cost-effectiveness of care.