Transforming Primary Care Payment in Hawaii

To design an innovative payment system that improves upon fee-for-service (FFS), incorporates behavioral economic principles, and improves work satisfaction among primary care physicians (PCPs) while improving quality and reducing health spending at the state level. Second, to test the incremental effectiveness of two additional interventions: (1) shared financial incentives between physicians and poorly controlled diabetes and (2) social comparisons ranking physicians on quality metric performance and total cost of care.

The goal of this project is to transform the Hawaii Medical Service Association (HMSA) primary care provider payment model to better incentive population health while bending the increasing trend of health spending in the state. Primary care and overall spending patterns will be studied to lay the foundation for a more rationally designed model. This model deliberately shifts away from FFS and includes three components: 1) a risk-adjusted per-member, per-month (PMPM) base payment, 2) an enhanced quality incentive program with larger bonus amounts and 3) a total cost of care incentive at the PO level. The aim is to build on the success of the Alternative Quality Contract (AQC) program implemented by Blue Cross Blue Shield (BCBS) of Massachusetts. While the AQC is used as a starting point, the study introduces and tests a number of innovations using concepts from behavioral economics. First, the move away from the FFS chassis to a PMPM-based capitated payment. Second, 20% of the PMPM payment is at-risk based on metrics designed to increase engagement between HMSA and physicians and engagement with performance feedback. Third, the number of metrics in the quality incentive program is drastically reduced from over 60 metrics to 10-12 per specialty. Fourth, the scoring of quality incentives incorporates rewards for improvement, rather than exclusively attainment of thresholds, to activate physicians along the entire performance distribution. In addition to implementing the new payment model, the initial experiment will include a test of two additional behavioral concepts: social comparisons for physicians and a shared incentive for physicians and poorly controlled diabetics tied to improve glycemic control.