Risk adjustment is commonly used by health insurers to increase the accuracy of cost models by taking into account the risk characteristics of different populations, usually by providing a risk score for each individual. Milliman consultants are commonly asked to perform risk adjustment and are always looking to advance its analytical utility. In a recent paper, Milliman consultant Ksenia Draaghtel demonstrates that
…segregating risk by service category better represents the differences in utilization and cost within each component, and is an important aspect in actuarial pricing. Inpatient, outpatient, physician, and pharmaceutical services possess different characteristics with respect to the utilization frequency, cost severity, speed of claim payment, and underlying trends.
The simulations used in the paper show how these techniques can increase profitability.
Risk adjustment has become more newsworthy recently with the resurgence of interest in provider cost-sharing as a mechanism of healthcare cost control. Two recent Milliman papers discuss the issue. The first, “Risk adjustment and its applications in global payments to providers” by Rong Yi, Jon Shreve, and Bill Bluhm, is a great place to start if you are new to the topic. It starts with a rationale for using risk adjustment in this field. Take two patients, one healthy, one with chronic diseases, each patient being treated by different providers. Assume that healthcare payments are determined by patient age and gender. Without risk adjustment, the provider with the sicker patient would appear overpaid and inefficient and would miss out on efficiency-based performance bonuses. The paper goes on to discuss risk adjustment methodologies, concurrent and prospective models, data issues, and other important considerations.
The second paper by the same authors, “Risk adjustment: Important considerations for global payments to providers” addresses specific issues in more depth including the need for models to be transparent and meaningful, dealing with imperfect coding, and ways to improve risk adjustment through better design, richer data, and more advanced modeling techniques.
Finally, a Milliman Healthcare Reform Briefing entitled “Controlling healthcare costs the old, new way” treats risk adjustment in the context of the history of provider cost sharing. It discusses the largely failed attempts at provider risk sharing in the 1990s and what needs to happen if the latest reform attempts are to be successful. Relevant to the discussion of risk adjustment, the authors say:
Risk adjustment was a hallmark of the prior move toward provider risk sharing. These early efforts evolved to become today’s risk-based performance measures and have been integral in developing new incentive systems.
The role for risk adjustment, however, is broader than this. In addition to calibrating payments under the new incentive systems, risk adjustment can also refine the measurement of provider performance and utilization of resources. Risk adjustment can identify opportunities for improving quality and utilization, serving as a way of recognizing and enacting opportunities for improvement.
Risk adjustment is an analytical tool with a wide and growing variety of important applications. Well-designed risk adjustment models are essential not only to insurer profitability, but to the success of key healthcare reform initiatives.