Diagnosis-based Payments in Regulated Insurance Markets
Over the last 20 years, there has been a proliferation of health plan and provider payment policies where payments are adjusted for health status (i.e. risk adjustment). Diagnoses from encounters between physicians and patients are typically used to construct risk scores that are then used to determine insurer or provider payment. The goal of these payments is to limit adverse selection problems, such as cream skimming and market unraveling, by paying more for sicker, higher cost individuals. The papers in this session examine these types of policies from two angles. The first two papers focus on how these policies incentivize insurers and providers to code their patients more aggressively. The first paper finds that Medicare Advantage plans act on this incentive and extract billions of dollars in overpayments from the government every year. The second paper identifies the characteristics of hospitals that are correlated with the hospital’s ability to respond quickly to a change in coding incentives in the Medicare program. The final paper studies how effective these diagnosis-based payments are at limiting the type of cream-skimming behavior they’re intended to prevent. This paper studies how the Medicare Part D prescription drug plan benefits change as the relative sizes of the diagnosis-based payments change, finding that as the payment tied to a particular health condition increases, cost-sharing for drugs used by individuals with that condition decreases, suggesting that insurers do adjust the benefits offered by their health plans in response to the diagnosis-based payments.