Testing the DMAA's recommendations for disease management program evaluation

The objective of this study was to compare and contrast findings regarding the financial savings projections of the disease management (DM) programs of 2 large employers based on different evaluation methods. In particular, this research tests the impact of differences in assumptions on the underlying growth rate of group health costs, exclusions of high-cost conditions and claims, and the length of the baseline period for determined health care costs. A pre-post study design was used. The data for this research came from 2 large employers in the consumer goods industry with comprehensive Health and Productivity Management programs. It contained medical and prescription drug claims and health plan enrollment data as well as program activity data from 2001 to 2005, covering an average yearly sample size of 201,037 members with 12 consecutive months of enrollment. Analyses were done on group-level averages using nominal cost data and were run to reflect the impact of a DM-only intervention. While the trend estimate and length of baseline had the largest effects on estimated program impact, the use of exclusions had an important effect as well. These findings demonstrate the importance of developing and instituting a standardized evaluation methodology. Without increasing consistency in the way evaluators develop their methodologies, it will remain difficult to be able to compare one evaluation to another, or to have faith in the results at hand.