Estimating Demand for Substance Use Disorder Treatment Models

A large and growing research literature has focused on whether collaborative care may improve detection and treatment of behavioral health conditions including substance use disorders (SUDs). However, very little work has been done to investigate the demand for collaborative care compared with other treatment models.  Given that only about 10 percent of individuals with SUD are currently in treatment, increasing treatment rates is an important goal.  Although significant research has examined strategies to increase the intensive margin of treatment, much less research has focused on the extensive margin.  To fill this gap, we conducted a nationally representative randomized internet-based experiment of individuals. First, individuals were screened for substance use disorders and current treatment status.  Respondents screening positive for SUD and not in treatment (N=1100) were randomized to one of three treatment vignettes describing key components of three care models: specialty treatment provided through an addiction treatment center, collaborative care including services provided by a care manager, and care provided by a primary care provider. Outcomes include self-reported willingness to enter treatment, willingness to pay for treatment at different (randomized) prices and, for those not willing to enter treatment, whether small payments increase respondents’ willingness to enter treatment.  We also examine heterogeneity in patients’ treatment preferences, and test whether having more treatment models available might increase overall treatment rates.   Early results examining a subset of the data suggest significantly more individuals are willing to enter treatment provided in primary care compared to specialty care settings (42 % versus 26 %).  Comparing primary care to collaborative care, we find no significant differences in self-reported willingness to enter treatment.  Using variation in price of treatment embedded in the survey, we will estimate the elasticity of demand for different SUD treatment services.  Preliminary analyses suggest that as expected as price increases, demand for all treatments declines. For example, compared to free care, when treatment was associated with a $30 out-of-pocket payment significantly fewer individuals were willing to enter treatment (34 % versus 16 %).  Among those not willing to enter treatment, providing small payments increased treatment rates.  If services were associated with a $5 payment, an additional 9 percent of respondents reported a willingness to enter treatment. The overarching aim of this study is to better understand how the broader availability of SUD treatment in different settings might affect demand for substance treatment. Data collection is complete as of November 2013.  Additional results will be available by the June 2013 conference.