THE IMPACT OF FINANCIAL INCENTIVES AND INFORMATION ON REFERRALS FOR SPECIALTY CARE: EVIDENCE FROM A PATIENT-CENTERED MEDICAL HOME PROGRAM

Because they decide where to refer a patient for specialty care, primary care physicians (PCPs) play a central role in determining overall health care cost and quality. While differences in quality are hard to measure, variation in cost across specialty-care providers and downstream decisions leading to hospitalizations can be substantial, and influence the total cost of care.

As part of a multi-year program evaluation, we examine the effectiveness of an intervention that provides PCPs with financial incentives and information about the cost and quality of specialty care for their attributed patients. The patient-centered medical home (PCMH) program was designed and implemented by CareFirst Blue-Cross Blue Shield of Maryland in 2011 and is expected to continue through 2016. In 2009, ten of the most frequently performed hospital-based surgeries accounted for 44% of claims payments, but only for 29% of all admissions. Across CareFirst’s market region in Maryland, Virginia, and the District of Columbia, the cost of total hip replacement varied by as much as 100%.

The PCMH program offers PCPs a 12% fee increase for signing up to participate in the program; annual outcome incentive awards that provide an additional fee increase based on a provider’s actual costs incurred relative to projected cost growth; care-coordination support for patients with multiple chronic conditions; and a web-based portal allows participating PCPs to track their patients’ total cost, identify lower-cost alternatives for specialty care, monitor high-risk patients who have multiple chronic conditions, and develop strategies for follow-up.

Using member-level claims data that span the pre- and post-implementation periods, we examine whether the intervention affected how often members received select referral-sensitive surgical procedures and where they received them. To identify the effect of the intervention, we assess the pre-post trends in referral patterns for patients who received care from participating PCPs to trends in referral patterns for patients in a comparison group of PCPs who did not participate in the PCMH program.