The Cost Effects of a Private Health Insurance Shared Savings Model

Shared savings models, where physician practices are paid a portion of the total savings achieved for their patients, have found increased popularity to create cost reduction incentives. These models have yielded modest effects in the public payer setting, but few studies have assessed the effects of private payer variations to the model. This paper assesses the cost effects of a large private payer-based model.

In this specific model, Accountable Care Organizations (ACOs) agree on a projected cost target with the payer. The practices continued to be paid under a fee-for-service arrangement, but if total costs are under the target, the savings are shared between the insurer and the ACOs. The launch was staggered and virtually all non-participating practices eventually participated the program. Using differences-in-differences and event study designs with several years of privately held healthcare claims data, I assess the effects of the program by comparing participating practices against non-participating practices while controlling for patient risk scores, various fixed effects, and including group-specific time trends.

Cost outcomes include overall, pharmacy, medical, inpatient, emergency room (ER), and preventative care costs. Efficiency measures include use of imaging, potentially preventable ER visits, and generic drug utilization rates. Utilization outcomes include inpatient, ER, and length of stay averages. Preliminary results suggest that the program changed utilization and cost patterns among the treatment practices.