The cost effectiveness of ESDM

Saturday, May 17, 2014: 11:54 AM
Marquis BC (Marriott Marquis Atlanta)
D. S. Mandell1, Z. Cidav2, J. Munson3, A. Estes4 and G. Dawson5, (1)Psychiatry, University of Pennsylvania School of Medicine, Philadelphia, PA, (2)Center for Mental Health Policy and Services Research, University of Pennsylvania, Philadelphia, PA, (3)University of Washington, Seattle, WA, (4)Speech and Hearing Sciences, University of Washington, Seattle, WA, (5)Psychiatry and Behavioral Sciences, Duke University, Durham, NC
Background:  A number of intervention programs for very young children have been developed that demonstrate remarkable efficacy in improving cognition, social interaction and adaptive behavior. These interventions are very expensive to implement, however.  Concerns have been raised about their cost effectiveness and place within the healthcare system and publicly funded education system. In response, a number of simulations have been published, suggesting that children who receive intensive early intervention will have lower societal costs, especially in their use of medical and educational services over the lifespan. None of these simulations has contained data from a cohort of children who have received these interventions, however, nor have those been compared with data from children who did not receive one of these evidence-based interventions.

Objectives:  to examine the cost effectiveness of one such intervention, the Early Start Denver Model (ESDM)

Methods:  Data were obtained from the original randomized controlled trial of ESDM. Service use data were collected on all children in the trial during the intervention period and for four years post intervention. Weekly service use was calculated in multiple categories and included the intervention itself (ESDM), early intensive behavioral intervention, functional therapies, special education, parent training, social skills training, general education, and other costs. Services were monetized by pooling data from previous studies. Costs were log transformed to normalize the distribution. T-tests were used to assess the statistical significance of differences in cost between the intervention and control groups.

Results:  Analyses are ongoing. Preliminary results suggest that during the intervention period, service costs for the intervention group averaged approximately $10,000 per child, compared with $5300 for the control group. Post intervention, service costs for the intervention group averaged approximately $4400 per child, compared with $5600 in the control group. Differences in cost post intervention were especially apparent in use of early intensive behavioral intervention and special education services. Return on investment occurred approximately 10.5 years post intervention (i.e., when most subjects would be entering high school). That is, accrued costs were equal between groups at that point, and were projected to be higher for the control group than for the intervention group after that point.

Conclusions:  The results suggest that the cost effectiveness of ESDM is realized much earlier in the lifespan that would be projected from prior simulation studies. Of note, the current study did not include many other direct and indirect costs whose inclusion could have greatly increased the observed difference between groups. Challenges faced in analyzing these data suggest the need for easy-to-complete service use and cost instruments that are included as part of intervention trials.