Effects of Mental Health Parity on High Utilizers of Services: Pre-Post Evidence From a Large, Self-Insured Employer

Using enrollee and dependent health insurance claims data provided by a large, self-insured regional employer, this study employed a pre-post design to compare changes in utilization of treatment for mental illness and substance use disorders following elimination of a 30-visit annual limit for mental health outpatient treatment, group therapy, and testing. Examination of the written policies regarding benefits and conversations with the health plan managers confirmed that these plan changes occurred in 2010 in response to the 2008 federal parity legislation and associated 2010 regulations implementing mental health and substance use benefits parity. They were the only notable changes related to mental health and substance use services at that time. In 2009, copays for the first 30 mental health visits annually rose from $15 to $20 per visit, consistent with copays for medical visits, with full cost charges for visits 31 and above. All other benefits remained consistent before and after the federal parity regulations. Visit data by treatment type, net insurer expenditure, enrollee copayment, and inpatient or outpatient services for each enrollee were calculated for 2009, the year before elimination of the 30-visit limit, and 2010, the year in which the limit was eliminated. This study was approved by the Institutional Review Board of the University of Michigan.

Analysis of claims data was limited to the 20,349 employees and 23,506 dependents (including spouses) who were continuously enrolled in the employer-sponsored health plan for the 24-month entirety of 2009 and 2010. Descriptive statistics were obtained, and a chi square test was used to compare service utilization across the two years. All analyses were conducted using SAS 9.3.

Enrollees were predominantly female (N=23,584, 54%), with a mean±SD age of 46.5±12.7 for subscribers and 24.7±19.4 for dependents; 50.6% (N=11,883) were less than 18 years of age, 28.2% (N=6,620) ages 18–44, 18.4% (N=4,314) ages 45–64, and 2.9% (N=689) age 65 and over.

From 2009 to 2010, the mean number of mental health visits per subscriber increased 12%, from 1.0±4.0 to 1.2±4.6, while the mean number of mental health visits per dependent increased 11%, from .8±3.4 to .9±3.8 (p<.001) (Table 1). The percentage of subscribers with at least one mental health visit increased from 13.7% to 14.5%, while the percentage of dependents with one visit increased from 11.4% to 12.3%. The number of mental health visits by subscribers ranged from zero to 44 in 2009 and from zero to 78 in 2010; the number of mental health visits by dependents ranged from zero to 46 in 2009 and from zero to 82 in 2010.

Spending by the health plan for mental health visits increased from $68.67 in 2009 to $74.88 in 2010 among subscribers with a mental health visit, a 9% increase; plan spending for visits by dependents with a mental health visit rose from $56.81 in 2009 to $62.81 in 2010, an increase of 11%. Among users of mental health services, out-of-pocket spending for co-pays rose from $15.17 in 2009 to $23.10 in 2010 for subscribers and from $11.75 in 2009 to $17.16 in 2010 for dependents, an increase of approximately 50% (p<.001).

The number of subscribers with 31 or more mental health visits increased by 255%, from 29 in 2009 to 103 in 2010, while the number of subscribers with exactly 30 visits decreased by 74%, from 53 to 14. There was a 176% increase, from 25 to 69, in the number of dependents with 31 or more mental health visits between 2009 and 2010, and a 64% decrease, from 28 to 10, in the number of dependents with exactly 30 visits (Table 1). Mental health parity had a statistically significant effect on the increase in the use of mental health services by high utilizers (p<.001, analyses available upon request). Thirty-seven (70%) subscribers who had exactly 30 mental health visits in 2009 had between one and 29 visits in 2010, and 12 (23%) had 31 or more visits. None of the dependents who had exactly 30 mental health visits in 2009 had exactly 30 visits in 2010; 24 (75%) had one to 29 visits, and four (14%) had 31 or more. Although high utilizers represented less than 1% of the health plan’s enrolled members, the increase in visits above the 30-visit threshold accounted for 16% of the increase in mental health costs (analyses available upon request).

Between 2009 and 2010, the average number of annual medical visits, excluding mental health visits, increased 1% for subscribers (3.75±4.1 versus 3.79±4.4) and decreased 7% for dependents (3.2±3.5 versus 3.0±3.4), and plan expenditures for medical care increased by 5% for subscribers and 16% for dependents.

Outpatient visits for a substance use disorder by subscribers increased by 20%, from 31 visits per 1,000 subscribers in 2009 to 37 visits per 1,000 subscribers in 2010. Visits by dependents for a substance use disorder increased by 8%, from 46 visits per 1,000 dependents in 2009 to 50 visits per 1,000 dependents in 2010. The total number of subscribers with at least one visit for a substance use disorder increased 3%, from 132 to 136, while the number of dependents with at least one visit for a substance use disorder decreased by 1.2%, from 171 to 169. However, no subscriber had 30 or more substance use disorder visits in 2009 and 2010, and only three dependents had 30 or more visits in 2009 compared with four in 2010. Subscriber total plan expenditures increased by 25%, however, and out-of-pocket spending by subscribers increased by approximately 40%. Despite these changes in utilization of services for substance use disorders, both plan expenditures and out-of-pocket spending for dependents decreased.

This study assessed the impact of parity legislation on treatment utilization among persons who exceeded previous limits on the use of mental health or substance use services. We found that expanded benefits were associated with increased treatment utilization among high utilizers of mental health treatment. These findings suggest that the expansion of mental health benefits as a result of the ACA, MHPAEA, and MHPA has the potential to increase treatment utilization by individuals who had previously met or exceeded the coverage limits of their health plan.

Our findings showed a significant increase in mental health services utilization for both subscribers and dependents after visit limits were removed as a result of the federal parity legislation and associated regulations. The number of subscribers and dependents with any mental health visits increased, as did plan expenditures for mental health care.

Mental health parity had a statistically significant increase on the use of mental health services by high utilizers. Upon removal of the 30-visit maximum in 2010, the number of enrollees with more than 30 mental health visits increased substantially. Because the share of enrollees with exactly 30 visits, the previous annual limit, decreased, it is likely that many persons whose utilization was previously curtailed by insurance limits were able to increase utilization to 31 or more visits per year. These results suggest that there was a large unmet need for mental health treatment among health plan enrollees who were already high utilizers prior to the expansion of benefits. Whereas Douven and colleagues noted “spikes” in data for the exact number of visits near the limit, the results here did not support that behavior (11).

The effects of parity on utilization of substance use services were less clear. The average number of visits for a substance use disorder increased by 20% for subscribers and 8% for dependents. Few enrollees were high utilizers of substance use disorder services, and they did not experience a significant increase in service utilization following the expansion of benefits. The limited growth in substance use services—in fact, a decreased number of visits among dependents—is difficult to interpret but may be due in part to subscriber concern that use of employer-sponsored insurance would alert their employers to their substance use (12).

This study had several limitations. The lack of a control group limited the ability to account for the effect of uncontrolled variables on use of mental health and substance use services—for example, the effects of the economic recession, internal operations of the health plan, and changes in attitudes regarding behavioral health care. Another limitation was the two-year study period, which may have been insufficient to capture the longer-term effects of parity on utilization of mental health and substance use services. Despite these limitations, our results suggest that national expansion of access to mental health and substance use benefits will result in increased use of care by patients who previously exceeded their benefit coverage.

The ACA, MHPAEA, and MHPA collectively have expanded insurance coverage for substance use disorders and mental health services. This study sought to determine the effect of the expansion of mental health and substance use benefits on the utilization of mental health and substance use services by high utilizers–health plan enrollees who had met or exceeded the maximum number of visits in the prior year. Our results showed that following the expansion of benefits, persons who used more services increased utilization of mental health benefits but not substance use benefits. The increase in use of mental health benefits was likely beneficial for this vulnerable segment of the population, but it did not have a large influence on the overall costs for the employee population.