Physician Response to Provider Networks
In a tiered physician network, physicians are grouped into at most three tiers based on price and/or quality, and consumers have financial incentives to use physicians in the preferred tier. The Massachusetts Group Insurance Commission (GIC), a quasi-state agency that administers health insurance benefits to over 400,000 state and municipal employees, their dependents, and retirees annually, has included a tiered physician network in all six of its non-retiree health plans since 2008. Previous work has documented that patients respond to tiered networks by substituting toward preferred-tier providers. Physicians also may be motivated to change their behavior (e.g., practice medicine according to prescribed quality metrics, alter resource use, lower prices, shift referrals) either to avoid loss of market share or due to intrinsic motivation to obtain a preferred tier-ranking. Physician response to tiered network benefit design is unknown.
We use a unique dataset containing physician-level performance data from 2013-16. Data include physician level efficiency scores (utilization and spending at the episode and procedure level) and quality scores (performance at the measure level) based on rolling three years of claims data across six health plan carriers and that cover over 2,000,000 lives. Individual physicians receive report detailing their performance across efficiency and quality measures annually; reports were also provided to practice management beginning in 2015. In this study we examine quality, efficiency, and price responses by physicians as a function of their incentives to be in a preferred tier. We break down physician response as a function of initial performance, level of competition, and patient composition. Results of this study will provide evidence as to how the strength of incentives in tiered physician networks affect individual provider behavior and the performance on quality and cost-efficiency metrics across the population of physicians included in the network.