Comparison of Risk Assessment Methods General Cons deratlons

There exist a number of general considerations beyond predictive accuracy when selecting a risk assessment method for use in risk adjustment. In particular, these issues relate to the practical side of this process, such as whether it can be implemented and administered, whether the data exist to support its application, and what types of incentives does it provide to perform risk selection, game the system, or to promote efficient health care. We identified four general criteria for comparing methods: 1. Practicality~administrative cost. A risk adjustment system cannot be so complex and costly that it cannot be applied under real life circumstances. It should be understandable and somewhat straightforward to explain and apply. Furthermore, the data required for the approach must be readily available or feasible to obtain or develop at a reasonable cost. 2. Ability to restrict manipulation. A system should limit the ability of health plans to benefit financially by "gaming" the system. The model should rely on data that are objective and easily validated. 3. Timeliness and Predictability. In setting premiums, health plans should be able to predict with some accuracy the amount of the carrier transfer and how it will affect their premiums. 4. Incentives for efficiency and quality care. A risk adjustment system should provide no disincentive for providing efficient and high-quality care. We compare each of the methods based on these principles below. A. Practicality/Administrative Cost