Behavioral healthcare services use in health savings accounts versus traditional health plans.

BACKGROUND: Numerous studies have examined behavioral health services via employer-sponsored health insurance cost-sharing measures. Their results clearly indicate that health plan design matters a great deal with respect to behavioral health utilization. It is also clear that there remain a number of unresolved issues, particularly with respect to the effects of a switch from traditional plan designs to high deductible, consumer-driven policies. Health Savings Accounts (HSA) have been well described in the literature with some comparisons to traditional healthcare plans, however no reports have been made about their use for behavioral health treatment.

AIMS: We sought to estimate the impact switching to a consumer driven health plan (CDHP) with a health savings account had upon the utilization of behavioral health care. Utilization of behavioral health services were reviewed from claims data over three years (2005 through 2007). Comparisons were made between members who switched from traditional health plans to consumer driven health plans in 2007 with health savings accounts and members who remained in traditional health plans.

METHODS: A pre-post study design was applied to two cohorts, stayers and switchers. The stayer cohort consisted of traditional health plan members enrolled from 2005 through 2007. Stayers were offered a health savings account in 2006 and 2007, but opted to remain in traditional health plans. The switcher cohort consisted of members enrolled in traditional plans in 2005 who opted to switch to a health savings account for two years thereafter (2006 and 2007). The use and intensity of behavioral health services in each study year were generated from claims data. Logistic and OLS regression analyses were applied to behavioral health services use and outpatient intensity measures respectively with independent variables post years, cohort and their interaction terms. Both analyses controlled for demographic variables. Additional behavioral disorder variables were added to the intensity regression.

RESULTS: Members who switched to a health savings account plan were slightly less likely to initiate behavioral health services in each post year relative to members who stayed in traditional health plans. Of those who sought outpatient behavioral services, there was no difference between cohorts in the intensity of behavioral health services they received.

DISCUSSION: Our results suggest enrollment in CDHPs moderately affects the use of behavioral health services but do not affect the intensity of outpatient behavioral health services conditioned on initiating these services. These finding are somewhat limited in that specific information about benefits were not included in the study. These results are also subject to self-selection bias. Members who switched to CDHP may be influenced to do so by other unknown factors that bear on their behavioral health.

IMPLICATIONS FOR FURTHER RESEARCH: Recent growth in the number of health savings accounts and current attention to mental health legislation warrant answers about behavioral health spending and efficacious utilization of behavioral health services. Further studies which include behavioral health services outcomes and quality of care gleaned from claims data can answer questions about the efficiency of health savings accounts.