Observational Intensity Bias Associated With Illness Adjustment

Cross Sectional Analysis of Insurance Claims

In a study led by the Dartmouth Atlas Project and The Dartmouth Institute for Health Policy & Clinical Practice, researchers raise significant questions about the risk adjustment that Medicare and others apply to insurance claims data in an effort to make fair, apples-to-apples comparisons about the performance of doctors and hospitals and fairly credit providers for treating patients who are sicker than average.

The study published in BMJ, examines commonly used risk-adjustment methods that incorporate patients’ diagnoses, and finds that these methods can cause regions and hospitals with more physician visits, referrals, tests, and imaging that make some patient populations appear to be sicker than others when they are not. As a result, these regions and providers with more diagnoses receive higher reimbursements.

The paper concludes that regions where patients see doctors more frequently, have patients that appear to be, but are not, sicker than other regions. It suggests that adjusting the data based on a region’s rate of patient visits to physicians provides a more accurate representation of the patient population.

Most Requested