Association for Public Policy Analysis & Management

Background: Historically, access to behavioral health (BH) services for children and adolescents has been fraught with challenges and barriers. The 2008 Mental Health Parity and Addiction Equity Act (MHPAEA) was the most recent federal attempt to address access barriers and decrease the financial burden associated with treating BH disorders by requiring parity in coverage between BH and physical health services. These policy efforts are especially important for children, yet the impact of MHPAEA on access to children’s BH services has yet to be explored. We use several linked administrative databases from Optum®, a fully owned subsidiary of UnitedHealth Group, to address this gap.

Objective: This study examines utilization and expenditures of specialty BH services for children with private insurance from 2008-2013 to explore possible associations of MHPAEA with child access to BH services (pre-implementation=2008-2009, transition=2010, post=2011-2013).

Sample: Children ages 4-17 in self-insured “carve-in” and “carve-out” plans from large employers. In “carve-ins,” BH and medical care are covered through the same insurance plan, whereas in “carve-outs,” BH and medical coverage are administered separately. The unit of analysis is the person-month (N=80,655,188).

Methods: This study links insurance claims, eligibility, plan and employer data and employs an interrupted time series model allowing for intercept and slope changes for the transition and post-parity periods. Outcomes included total, plan and patient out-of-pocket (OOP) expenditures; outpatient assessment, individual and family psychotherapy and pharmacotherapy visits; and structured outpatient, day treatment, residential and inpatient days. Outcomes are estimated using two-part regression models distinguishing the probability of a positive outcome (e.g., any expenditures) from the conditional level of the outcome (e.g., amount of expenditures among those with any). Generalized Estimating Equations were used to account for clustering and p≤.01 was used as the significance cutoff.

Results: For carve-in enrollees, there were significant increases in the levels and/or slopes of total and plan expenditures post-parity due to increases in both the percent of children with any expenditure as well as higher expenditures among service users. To illustrate the magnitude of these effects, in July 2012, mean per-member-per-month total expenditures were predicted to be $5.65 without parity but $8.72 with parity. Neither penetration rates nor conditional averages changed significantly for patient OOP costs. Significant overall increases were seen for utilization of most outpatient services (assessment, individual and family psychotherapy, pharmacotherapy). Among carve-out enrollees however, estimates were largely insignificant and where significant, were mixed in sign.

Conclusions: The study detected more changes in expenditures and utilization rates for carve-in than carve-out enrollees, which may be partially explained by lower baseline coverage among carve-in enrollees. No clear pattern of significant changes emerged for carve-out enrollees. In comparison, for carve-in enrollees, the introduction of MHPAEA led to significant increases in utilization of outpatient services and an overall increase in plan expenditures without a parallel increase in patient expenditures. This suggests that for plans that did not already offer generous BH benefits, the introduction of MHPAEA was associated with an increase in access for children without a commensurate increase in financial burden for families.