An ACA Provision Increased Treatment For Young Adults With Possible Mental Illnesses Relative To Comparison Group

The Affordable Care Act (ACA) has been considered a transformative policy for increasing access to care for people with mental health and substance use disorders. ¹ Although the largest ACA coverage expansions began in January 2014, one important early provision has allowed young adults to remain on their parents’ health insurance as dependents. The dependent coverage provision went into effect in September 2010. It applies to people ages 19–25 whether or not they reside outside of their parents’ home, are their parents’ tax dependents, or are married. ²

The provision initially did not require insurers to cover people who already had an offer of employer-sponsored insurance, but this exemption was phased out in 2014. ² Early estimates suggest the provision increased the number of insured people in the target population by over two million ^3–5 and increased the use of emergency department and hospital services. ^6,7

The dependent coverage provision could be especially important for young adults because mental health and substance use disorders peak in young adulthood, ⁸ and insurance coverage has historically been low among young adults. ⁹ At least two studies have examined the impacts of the ACA provision. Yaa Akosa Antwi and coauthors ⁶ used data from the National Inpatient Sample, a national sample of hospital admissions, and found that the provision had both increased inpatient admissions overall among young adults and increased the fraction of young adults admitted with private insurance. The same study found that mental health admissions had increased by 5.5 percent, relative to a comparison group of older adults, and that mental health admissions increased more than all non-pregnancy-related admissions (including non–mental health admissions). ⁶

In the second study, Paul Fronstin ¹⁰ used claims data from a single large employer to compare young adults who were covered as employees and those who were newly covered as dependents. He found that people enrolled through the new provision were higher users of mental health and substance abuse treatment than people in the same age group with coverage in their own name.

Other recent insurance expansions that were not specific to young adults provide additional evidence about the potential impacts of access to health insurance on mental health and substance abuse treatment. A natural experiment in Oregon in which low-income adults were randomly offered the chance to enroll in public insurance found that the use of mental health treatment surged among adults who received an opportunity to enroll and that expanded coverage lowered the burden of depression symptoms. ¹¹ However, on average the Oregon sample was both poorer and older than the people affected by the ACA dependent coverage provision.

In contrast, Ellen Meara and coauthors found no change in hospital-based care for behavioral health among people ages 19–25 in Massachusetts after the state’s 2006 health reform. ¹² However, the reform increased financial protection for people with behavioral health diagnoses.

We extend the recent literature by using data from the National Survey of Drug Use and Health. This allowed us to estimate the impacts of the ACA dependent coverage provision on treatment in a nationally representative population of young adults who screened positive for possible mental health or substance use disorders.

We hypothesized that the provision would increase the use of both mental health and substance abuse services among people with likely disorders. In addition to increasing use overall, we expected that increasing private coverage rates would also lower out-of-pocket spending for people already in treatment—an effect that has been identified in previous insurance expansions and mental health coverage provisions. ^13,14

The ACA dependent coverage provision could also change the setting and type of care received by people needing treatment, especially among the previously uninsured. Uninsured people typically rely upon community clinics and other safety-net providers. ^15,16 In contrast, it may be relatively difficult for the uninsured to access specialist and residential treatment for substance use disorders and individualized psychotherapy for mental disorders. ¹⁷ Gaining private coverage could shift people toward specialty services and settings.

Study Data And Methods

Data

We combined data files from the 2008–12 rounds of the National Survey of Drug Use and Health, an annual cross-sectional survey of noninstitutionalized people that is administered by the Substance Abuse and Mental Health Service Administration. The survey covers the use and abuse of drugs and alcohol, mental health problems, and experiences with mental health or substance abuse treatment. Survey subjects are selected through a stratified random sample of addresses and answer questions using audio computer-assisted self-interviewing. To provide nationally representative estimates, we applied survey weights that adjusted for differential sampling and nonresponse rates.

Our main group of interest was people ages 19–25 who were likely to have either mental health or substance use disorders. However, because the survey reports age in categories, we could not exclude eighteen-year-olds. Therefore, our analysis group was people ages 18–25. There were 13,897 people in the mental health sample and 14,705 in the substance abuse sample (5,585 people met the criteria for and were included in both groups). Our comparison group was people ages 26–35 with the same symptoms.

We categorized people as possibly needing mental health treatment using a measure of serious psychological distress in the previous year. The measure was based on the Kessler-6 inventory, a six-item scale designed to be used by lay interviewers that includes questions about symptoms of depression and anxiety. ¹⁸ We categorized people as possibly needing substance abuse treatment if they met screening criteria for either dependence on or abuse of alcohol or drugs (not including tobacco).

Screening measures of substance abuse or dependence in the National Survey of Drug Use and Health are based on the Diagnostic and Statistical Manual of Mental Disorders , Fourth Edition (DSM-IV). They capture personal, legal, and school or employment problems related to substance use as well as problems in reducing use of the substance. For example, if a respondent indicated that she drank alcohol, she would be asked questions about alcohol abuse such as, “During the past twelve months did you have any problems with family or friends that were probably caused by your drinking?” She would also be asked questions about alcohol dependence such as symptoms of anxiety and nausea after cutting back on alcohol use. The survey measures of mental health and substance use disorders have been validated for screening in community-based samples of adults. ^18,19

We considered four outcomes separately for mental health and drug or alcohol treatment: the use of any treatment in the previous twelve months; if treatment was received, whether previous treatment had been paid for by private insurance (instead of some other source); the setting or type of previous treatment; and the coverage of mental health and substance abuse treatment among people with private insurance.

In a descriptive analysis, we examined differences in several variables that were likely to predict use independent of the ACA policy: sex, self-reported general health, marital status, race or ethnicity, employment status, receipt of any government assistance, and residence in a metropolitan area.

Statistical Analysis

Following other studies that have looked at the ACA dependent coverage provision, we used difference-in-differences regression to estimate the impact of the provision. ^4,5,20 This method identified the effect of the policy by comparing how much outcomes changed during the study period for people ages 18–25, relative to changes for people ages 26–35.

We considered people in the older age group as a counterfactual for trends in access and treatment in the absence of the policy. This is because the older people were assumed to be unaffected by the dependent coverage policy but affected by other time-varying factors, such as changes in the labor market or in access to substances.

To capture the average impacts of the policy, we used linear probability models. All models included indicators for age group and for whether the person was observed in the post period, after the implementation of the policy.

Each estimate is the interaction between the post period and the group of people ages 18–25 from a separate regression. For example, a positive 0.10 difference-in-differences estimate for the use of mental health treatment would signify that changes during the post period in the rate of mental health treatment for people ages 18–25 was 10 percentage points greater than the changes among people ages 26–35. In turn, this suggests that the ACA dependent coverage provision has had a positive effect on mental health treatment.

In our preferred specification we excluded data from 2010, since the provision was implemented in September of that year. However, our estimates were consistent when we treated 2010 as a year in the pre period. We also included a linear time trend. ²¹

Our preferred specification included covariates only for race or ethnicity and for residence in a metropolitan area. In the online Appendix ²² we report on regressions that included controls for self-rated health, employment status, receipt of social services, income, education, and marital status.

It is important to note that because differences were measured relative to the comparison sample, a positive difference-in-differences finding could occur either because an increase in the outcome over the baseline among people ages 18–25 was larger than the increase among people ages 26–35 or because there was no change (or a smaller decline) relative to the baseline among the younger group during a period of decline among the older group.

To display differences and difference-in-differences, we used predictive margins. ²³ Here we present model-predicted rates for each age group (18–25 and 26–35) by time period (the pre and post periods) after adjustment for covariates.

Models were weighted to be representative of the US population. Standard errors and confidence intervals accounted for the stratified sampling design.

Limitations

Some limitations should be considered. First, we could not measure changes in dependent coverage versus changes in other types of private coverage, since the National Survey of Drug Use and Health does not determine whether private coverage is provided directly to the person or as a dependent on someone else’s plan.

Second, our study design allowed us to control for time-varying factors that might have affected coverage and access to care across different age groups. However, we could not rule out the influence of other simultaneously occurring factors that might have had a disproportionate impact on young adults. For example, young adults were particularly affected by the economic downturn that occurred after December 2007. ²⁴ Nonetheless, our models were robust to controls for individual employment status.

Third, we could not determine whether or not some effects were connected to the early impact of federal mental health parity.

Fourth, for reasons of data confidentiality we could not examine age as a continuous variable. Also, because of the age categories in the National Survey of Drug Use and Health, we could not exclude eighteen-year-olds from our treatment group. This might have caused us to understate the true effects of the ACA dependent coverage provision.

Fifth, we could not examine detailed geographical differences, which might be pertinent because thirty-seven states had some form of extended dependent coverage provision before 2010. ²⁵ However, most of those provisions were more restrictive than the one in the ACA that superseded them, and there is little evidence that the state provisions increased coverage. ^26,27

Sixth, because the survey respondents reported on outcomes spanning the previous twelve months, some of the respondents from 2011 might have reported outcomes for a period when the extended dependent coverage policy had not been fully implemented. This would also tend to understate the impact of the policy.

Seventh, because we focused on subgroups with possible illness, our study did not account for effects of the provision among people who experienced recovery. We did not observe any population-level changes in the prevalence of these conditions during the study period. However, such changes might occur in the long term.

Finally, as is the case with all surveys that measure stigmatized conditions and behaviors, there is possible nonresponse and social desirability bias in the results of the National Survey of Drug Use and Health. The surveyors mitigate these factors by using audio computer-assisted self-interviewing and survey instruments that are designed to reduce social desirability bias. ²⁸

Study Results

Sample Statistics

Exhibit 1 presents characteristics of the samples with possible mental health and substance use disorders during the period before the provision was implemented (2008–09). Women made up 62.0 percent of the mental health sample, compared to 38.5 percent in the substance use sample. Non-Hispanic whites accounted for about two-thirds of both samples.

The largest segments in both groups were employed full time, but large fractions were not in the labor force (including students) or were unemployed. Just under half of the people in both groups had completed or spent some time in college.

About one-quarter of the members of both groups had incomes below the federal poverty level. Uninsurance was high in this period before implementation of the ACA dependent coverage provision: 28.7 percent of people in the mental health sample and 29.5 percent in the substance use sample were uninsured.

Sizable minorities in both groups rated their health below very good or excellent. In addition, 36.1 percent of the mental health sample also had possible substance use disorders, and 31.1 percent of the substance use sample also had possible mental disorders. More than half of the people in both groups resided in a large metropolitan area.

Unadjusted Trends

From 2008 to 2012 the use of mental health treatment by people ages 18–25 with possible mental health problems steadily increased, from less than 30 percent to about 33 percent ( Exhibit 2 ). In contrast, for people ages 26–35 the highest rate (above 47 percent) was in 2009, after which it declined to about 40 percent in 2011. The use of substance abuse treatment by people with possible disorders remained relatively flat for both age groups during the study period.

SOURCE Authors’ analysis of data from the 2008–12 National Survey of Drug Use and Health. NOTE The exhibit shows reported treatment in the previous year for mental health (MH) or substance abuse (SA).

People ages 18–25 with possible mental health disorders who used any treatment in the previous year experienced a slight increase in having their last treatment paid for by private insurance from 2008 (about 30 percent in) to 2012 (34 percent) ( Exhibit 3 ). However, the situation varied much more for people ages 26–35, with the rate of private insurance payment plummeting from about half in 2010 to about a quarter in 2012. The patterns also differed for people in the two age groups who received substance abuse treatment.

SOURCE Authors’ analysis of data from the 2008–12 National Survey of Drug Use and Health. NOTE The exhibit shows reported payments by private insurance for the last episode of mental health treatment and any payments within the previous twelve months by private insurance for substance abuse treatment.

Mental Health Treatment Estimates

After we adjusted for race or ethnicity, sex, and residence in a metropolitan area, we estimated that before implementation of the ACA dependent coverage provision (2008–09), 30.70 percent of people ages 18–25 with possible mental disorders had received any mental health treatment in the previous twelve months ( Exhibit 4 ). The comparable figure for people ages 26–35 was 44.97 percent.

The older group’s service use declined by 3.16 percentage points in the post period (2011–12), whereas the younger group’s service use increased by 2.16 percentage points ( Exhibit 4 ). Comparing trends in the two groups provided the main difference-in-differences estimate: a 5.32-percentage-point increase for the younger group compared to the older one after the ACA dependent coverage provision was implemented. This is equivalent to a 17.3 percent increase over the baseline rate (5.32 percentage points divided by $30.70\mathrm{percent}=17.3\mathrm{percent}$ ).

Compared to people ages 26–35, those ages 18–25 who used mental health treatment in the previous year were significantly more likely after implementation of the ACA dependent coverage provision to have reported that their most recent treatment was paid for by private insurance (12.94 percentage points) and significantly less likely to report that treatment was received without health insurance payment (−12.43 percentage points) ( Exhibit 4 ). These changes were very large, considering that 29.44 percent of the younger group at baseline said that their last treatment had been paid for by private insurance, and 59.88 percent reported that treatment was received without health insurance payment.

In contrast to our hypothesis, there were no significant changes among people ages 18–25 relative to people ages 26–35 using mental health treatment in the type or setting of treatment in the previous twelve months or in the percentage of people with private insurance who reported that their coverage included mental health treatment.

Substance Abuse Treatment Estimates

Compared to people receiving mental health treatment, an even smaller fraction (8.84 percent) of those ages 18–25 with possible substance use disorders received any treatment in the previous year before implementation of the ACA dependent coverage provision ( Exhibit 5 ). We found no significant changes in treatment use after the provision’s implementation between the younger and older groups in our study.

As with mental health treatment, we examined payer sources for substance abuse treatment in the previous year (respondents could report multiple payment sources for substance abuse, however). The coefficients for these estimates were in the hypothesized direction: The trends were toward an increase in payment from both private insurance and Medicaid and toward less uninsured care among people ages 18–25, relative to those ages 26–35 ( Exhibit 5 ). However, these estimates were not significant (estimates were imprecise because of smaller sample sizes using treatment). There were no significant changes in settings for substance abuse treatment in the previous year for the younger group relative to the older group.

Finally, we examined changes in reported coverage of alcohol and drug treatment among the people in our sample with private insurance. Before implementation of the ACA dependent coverage provision, 64.83 percent of younger adults in the sample said that their coverage included alcohol treatment, and 65.09 percent said it included drug treatment (we excluded people who responded “don’t know”) ( Exhibit 5 ).

During the study period the percentages of the younger adults reporting coverage of alcohol and drug treatment remained relatively constant, but the percentages declined substantially for the older adults. Thus, after implementation of the ACA dependent coverage provision, coverage of alcohol treatment was 13.01 percentage points higher among the younger group than before implementation, relative to the older group, and coverage of drug treatment was 12.68 percentage points higher.

Full-Sample Estimates

In addition, we examined regressions using the full population—that is, not only people with possible mental health or substance use disorders. These regressions illuminated the potential impacts of the provision on people with less severe symptoms who may still receive treatment. They also captured any changes in the burden of symptoms related to the implementation of the provision. These results are displayed in the online Appendix, ²² along with other additional sensitivity analyses.

In the full sample, the use of mental health treatment increased by 1.30 percentage points (95% CI: −0.01, 2.62) after 2010 among people ages 18–25, relative to those ages 26–35. On an absolute scale, this is a much smaller change than the 5.32-percentage-point reduction in the subsample with possible mental health problems. However, it represents a 12.1 percent relative increase over the baseline treatment rate of 10.8 percent (in comparison, the relative change for the subsample with mental health problems was 17.3 percent, as explained above).

Full-sample estimates for mental health payer source and treatment setting or type were also similar to the subgroup estimates. There were no significant changes after implementation of the provision in substance abuse treatment use, payer source, or setting in the full sample among the younger group, compared to the older group.

Discussion

We examined the early impacts of the ACA dependent coverage provision on mental health and substance abuse treatment among young adults. Our study had four main findings.

First, the provision substantially increased the use of mental health treatment among people ages 18–25 with possible mental illnesses, relative to similar people ages 26–35. Second, there was a 12.43-percentage-point decline in uninsured mental health visits among the younger age group relative to the older one and a commensurate increase in visits paid for by private insurance, as well as an increase in visits paid for by Medicaid. Third, there was no significant change in the use of substance abuse treatment overall. Fourth, more people ages 18–25 who had private insurance reported that their plan covered drug and alcohol treatment after implementation of the provision, relative to people ages 26–35.

Our findings related to mental health treatment are consistent with other work suggesting that private insurance became a greater payer for these treatments among young adults after the ACA dependent coverage provision ^4,5,29 and that mental health treatment increased in that age group relative to others. ^6,10 As findings from the Oregon study suggest, ¹¹ there may be pent-up demand for mental health treatment among the uninsured, who frequently do not seek treatment because of cost concerns. ³⁰

It is also important to consider the context of the mental health findings. Some estimated changes among young adults occurred because of relative, instead of absolute, increases in the outcomes. For example, the use of any mental health services among people ages 18–25 increased only slightly during the study period, while the same outcome decreased among people ages 26–35.

Caution is also warranted because our study period was relatively short. In addition, despite significant findings, some outcomes for people ages 18–25 did not follow a consistent trend from year to year in unadjusted analysis.

The study period included the aftermath of an economic downturn that disproportionately affected young adults. ²⁴ The ACA dependent coverage provision potentially smoothed out disruptions in access to care that might otherwise have occurred among young adults. That said, effects might have been different during a period of economic expansion.

The dependent coverage provision shifted mental health treatment costs for young adults to private insurers. However, the increase in premiums to employers and other enrollees are likely to have been relatively low, since relatively few young adults use these services intensively. For example, at one large employer, expenses for adult dependent coverage enrollees were higher than those for other comparable adults. Nonetheless, the group enrolling through the ACA dependent coverage provision increased overall expenditures by only 0.2 percent. ¹⁰ Before implementation of the provision, the Centers for Medicare and Medicaid Services estimated that it would increase premiums by about 0.7 percent. ³¹

In contrast to mental health treatment, we found that there was no relative increase in the use of substance abuse treatment. Previous work focusing on state substance abuse parity found that the policy—which lowered the cost of treatment—also increased admissions. ³² However, other studies have found that policies that reduce the cost of substance abuse treatment have a larger effect on shifting payer type for people already in treatment than they do on increasing access. ³³

One potential explanation is that expanded coverage, on its own, may be insufficient to increase demand for substance abuse treatment. Experience with the 2006 health reform in Massachusetts suggests that barriers such as copayments may continue to deter the use of services among those needing treatment. ³⁴ Noneconomic barriers such as stigma are widespread ³⁵ and may deter newly insured young adults from seeking substance abuse treatment. A limited supply of substance abuse treatment providers and waiting lists to enter treatment could also deter some newly insured people from seeking care. ³⁶

In addition to increasing coverage for previously uninsured people, the ACA dependent coverage provision may also have affected young adults who already had private insurance in their own name but chose to switch to their parents’ plans to take advantage of more comprehensive benefits. ³ This is consistent with our finding that mental health and substance abuse coverage increased during the study period among young adults relative to the comparison group. Substance abuse benefits in particular have historically been restricted in many private insurance plans. ³⁷ In our sample, fewer than two-thirds of the respondents ages 18–25 believed that their coverage included alcohol or drug treatment before implementation of the ACA provision.

The Mental Health Parity and Addiction Equity Act of 2008—which requires most private insurers to offer behavioral health benefits that are comparable to physical health benefits—is another major piece of federal legislation. It could conceivably have affected young adults disproportionately. However, reports indicate that implementation of parity was slow, ³⁸ and rules were not finalized until 2013. ³⁹

The federal parity law is one example of how the dependent coverage provision must be considered in the context of broader health system changes. In 2014 the ACA made new Medicaid funding available to cover poor adults—an expansion that roughly half the states have opted to adopt thus far. Additionally, coverage through new health insurance Marketplaces was implemented nationally. In combination, new coverage provisions are expected to increase the number of insured people by twenty-five million. ⁴⁰

One possibility is that new health insurance options, coupled with more publicity about receiving care, will increase the awareness of treatment options and the demand for mental health and substance abuse treatment, particularly among people who have not previously used such treatment. However, provider capacity in areas with higher demand could be strained, which would limit access. ⁴¹

Conclusion

The ACA dependent coverage provision appears to be a stepping-stone toward increasing mental health treatment among young adults with possible mental health problems. The act may also improve the comprehensiveness of substance abuse coverage for young adults with substance use disorders.

The long-term success of this provision will depend on whether or not it links young adults needing treatment to providers in their communities and ensures continuity of treatment. Enabling people to remain in the care of their current providers may help ease transitions across insurance plans. Young adults with treatment needs are also likely to benefit from initiatives to increase chronic disease management and integrate physical health care with other social and behavioral health services.

ACKNOWLEDGMENTS

Brendan Saloner gratefully acknowledges funding support from the Robert Wood Johnson Foundation Health and Society Scholars Program. Benjamin Lê Cook received funding support from the Agency for Healthcare Research and Quality (Grant No. 1R01HS021486).

NOTES

• 1 Mechanic D . Seizing opportunities under the Affordable Care Act for transforming the mental and behavioral health system . Health Aff (Millwood) . 2012 ; 31 ( 2 ): 376 – 82 . Go to the article, Google Scholar
• 2 Goldman TR . Progress report: the Affordable Care Act’s extended dependent coverage provision . Health Affairs Blog [blog on the Internet]. 2013 Dec 16 [cited 2014 Jun 23 ]. Available from: http://healthaffairs.org/blog/2013/12/16/progress-report-the-affordable-care-acts-extended-dependent-coverage-provision/ Google Scholar
• 3 Akosa Antwi Y , Moriya AS , Simon K . Effects of federal policy to insure young adults: evidence from the 2010 Affordable Care Act dependent coverage mandate . Cambridge (MA) : National Bureau of Economic Research ; 2012 Jun . (NBER Working Paper No. 18200). Google Scholar
• 4 Sommers BD , Kronick R . The Affordable Care Act and insurance coverage for young adults . JAMA . 2012 ; 307 ( 9 ): 913 – 4 . Crossref, Medline, Google Scholar
• 5 Sommers BD , Buchmueller T , Decker SL , Carey C , Kronick R . The Affordable Care Act has led to significant gains in health insurance and access to care for young adults . Health Aff (Millwood) . 2013 ; 32 ( 1 ): 165 – 74 . Go to the article, Google Scholar
• 6 Akosa Antwi Y , Moriya AS , Simon K . Access to health insurance and the use of inpatient medical care: evidence from the Affordable Care Act young adult mandate . Cambridge (MA) : National Bureau of Economic Research ; 2014 Jun . (NBER Working Paper No. 20202). Google Scholar
• 7 Mulcahy A , Harris K , Finegold K , Kellermann A , Edelman L , Sommers BD . Insurance coverage of emergency care for young adults under health reform . N Engl J Med . 2013 ; 368 ( 22 ): 2105 – 12 . Crossref, Medline, Google Scholar
• 8 Kessler RC , McGonagle KA , Zhao S , Nelson CB , Hughes M , Eshleman S , et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey . Arch Gen Psychiatry . 1994 ; 51 ( 1 ): 8 – 19 . Crossref, Medline, Google Scholar
• 9 Collins S , Robertson R , Garber T , Doty M . Insuring the future: current trends in health coverage and the effects of implementing the Affordable Care Act . New York (NY) : Commonwealth Fund ; 2013 . Google Scholar
• 10 Fronstin P . Mental health, substance abuse, and pregnancy: health spending following the PPACA adult-dependent mandate [Internet]. Washington (DC) : Employee Benefit Research Institute ; 2013 Apr [cited 2014 Jun 23 ]. (Issue Brief No. 385). Available from: http://www.ebri.org/pdf/briefspdf/EBRI_IB_04-13.No385.ADM.pdf Google Scholar
• 11 Finkelstein A , Taubman S , Wright B , Bernstein M , Gruber J , Newhouse JP , et al. The Oregon health insurance experiment: evidence from the first year . Q J Econ . 2012 ; 127 ( 3 ): 1057 – 106 . Crossref, Medline, Google Scholar
• 12 Meara E , Golberstein E , Zaha R , Greenfield SF , Beardslee WR , Busch SH . Use of hospital-based services among young adults with behavioral health diagnoses before and after health insurance expansions . JAMA Psychiatry . 2014 ; 71 ( 4 ): 404 – 11 . Crossref, Medline, Google Scholar
• 13 Barry CL , Busch SH . Do state parity laws reduce the financial burden on families of children with mental health care needs? Health Serv Res . 2007 ; 42 ( 3 Pt 1 ): 1061 – 84 . Crossref, Medline, Google Scholar
• 14 Azzone V , Frank RG , Normand SL , Burnam MA . Effect of insurance parity on substance abuse treatment . Psychiatr Serv . 2011 ; 62 ( 2 ): 129 – 34 . Crossref, Medline, Google Scholar
• 15 Wu LT , Ringwalt C . Use of substance abuse services by young uninsured American adults . Psychiatr Serv . 2005 ; 56 ( 8 ): 946 – 53 . Crossref, Medline, Google Scholar
• 16 Wells R , Morrissey JP , Lee IH , Radford A . Trends in behavioral health care service provision by community health centers, 1998–2007 . Psychiatr Serv . 2010 ; 61 ( 8 ): 759 – 64 . Crossref, Medline, Google Scholar
• 17 McConnell KJ , Gast SH , McFarland BH . The effect of comprehensive behavioral health parity on choice of provider . Med Care . 2012 ; 50 ( 6 ): 527 – 33 . Crossref, Medline, Google Scholar
• 18 Kessler RC , Barker PR , Colpe LJ , Epstein JF , Gfroerer JC , Hiripi E , et al. Screening for serious mental illness in the general population . Arch Gen Psychiatry . 2003 ; 60 ( 2 ): 184 – 9 . Crossref, Medline, Google Scholar
• 19 Jordan BK , Karg RS , Batts KR , Epstein JF , Wiesen C . A clinical validation of the National Survey on Drug Use and Health assessment of substance use disorders . Addict Behav . 2008 ; 33 ( 6 ): 782 – 98 . Crossref, Medline, Google Scholar
• 20 Cantor JC , Monheit AC , DeLia D , Lloyd K . Early impact of the Affordable Care Act on health insurance coverage of young adults . Health Serv Res . 2012 ; 47 ( 5 ): 1773 – 90 . Crossref, Medline, Google Scholar
• 21 Bertrand M , Duflo E , Mullainathan S . How much should we trust differences-in-differences estimates? Q J Econ . 2004 ; 119 ( 1 ): 249 – 75 . Crossref, Google Scholar
• 22 To access the Appendix, click on the Appendix link in the box to the right of the article online.
• 23 Graubard BI , Korn EL . Predictive margins with survey data . Biometrics . 1999 ; 55 ( 2 ): 652 – 9 . Crossref, Medline, Google Scholar
• 24 Burgard SA , Ailshire JA , Kalousova L . The great recession and health: people, populations, and disparities . Ann Am Acad Pol Soc Sci . 2013 ; 650 ( 1 ): 194 – 213 . Crossref, Google Scholar
• 25 National Conference of State Legislatures . Covering young adults through their parents’ or guardians’ health policy [Internet]. Denver (CO) : NCSL ; 2010 Sep 23 [cited 2014 Jun 23 ]. Available from: http://www.ncsl.org/research/health/dependent-health-coverage-state-implementation.aspx#State_Actions Google Scholar
• 26 O’Hara B , Brault MW . The disparate impact of the ACA-dependent expansion across population subgroups . Health Serv Res . 2013 ; 48 ( 5 ): 1581 – 92 . Medline, Google Scholar
• 27 Monheit AC , Cantor JC , DeLia D , Belloff D . How have state policies to expand dependent coverage affected the health insurance status of young adults? Health Serv Res . 2011 ; 46 ( 1 Pt 2 ): 251 – 67 . Crossref, Medline, Google Scholar
• 28 Turner CF , Villarroel MA , Rogers SM , Eggleston E , Ganapathi L , Roman AM , et al. Reducing bias in telephone survey estimates of the prevalence of drug use: a randomized trial of telephone audio-CASI . Addiction . 2005 ; 100 ( 10 ): 1432 – 44 . Crossref, Medline, Google Scholar
• 29 Kirzinger WK , Cohen RA , Gindi RM . Trends in insurance coverage and source of private coverage among young adults aged 19–25: United States, 2008–2012 [Internet]. Hyattsville (MD) : National Center for Health Statistics ; 2013 Dec [cited 2014 Jun 24 ]. Available from: http://www.cdc.gov/nchs/data/databriefs/db137.htm Google Scholar
• 30 Wang PS , Lane M , Olfson M , Pincus HA , Wells KB , Kessler RC . Twelve-month use of mental health services in the United States: results from the National Comorbidity Survey Replication . Arch Gen Psychiatry . 2005 ; 62 ( 6 ): 629 – 40 . Crossref, Medline, Google Scholar
• 31 CMS.gov . Center for Consumer Information and Insurance Oversight: young adults and the Affordable Care Act: protecting young adults and eliminating burdens on businesses and families [Internet]. Baltimore (MD) : Centers for Medicare and Medicaid Services ; [cited 2014 Jun 24 ]. Available from: http://www.cms.gov/CCIIO/Resources/Files/adult_child_faq.html Google Scholar
• 32 Wen H , Cummings JR , Hockenberry JM , Gaydos LM , Druss BG . State parity laws and access to treatment for substance use disorder in the United States: implications for federal parity legislation . JAMA Psychiatry . 2013 ; 70 ( 12 ): 1355 – 62 . Crossref, Medline, Google Scholar
• 33 McConnell KJ , Ridgely MS , McCarty D . What Oregon’s parity law can tell us about the federal Mental Health Parity and Addiction Equity Act and spending on substance abuse treatment services . Drug Alcohol Depend . 2012 ; 124 ( 3 ): 340 – 6 . Crossref, Medline, Google Scholar
• 34 Capoccia VA , Grazier KL , Toal C , Ford JH , Gustafson DH . Massachusetts’s experience suggests coverage alone is insufficient to increase addiction disorders treatment . Health Aff (Millwood) . 2012 ; 31 ( 5 ): 1000 – 8 . Go to the article, Google Scholar
• 35 Fortney J , Mukherjee S , Curran G , Fortney S , Han X , Booth BM . Factors associated with perceived stigma for alcohol use and treatment among at-risk drinkers . J Behav Health Serv Res . 2004 ; 31 ( 4 ): 418 – 29 . Crossref, Medline, Google Scholar
• 36 Hoge MA , Stuart GW , Morris J , Flaherty MT , Paris M , Goplerud E . Mental health and addiction workforce development: federal leadership is needed to address the growing crisis . Health Aff (Millwood) . 2013 ; 32 ( 11 ): 2005 – 12 . Go to the article, Google Scholar
• 37 Gabel JR , Whitmore H , Pickreign JD , Levit KR , Coffey RM , Vandivort-Warren R . Substance abuse benefits: still limited after all these years . Health Aff (Millwood) . 2007 ; 26 ( 4 ): w474 – 82 . DOI: 10.1377/hlthaff.26.4.w474 . Go to the article, Google Scholar
• 38 Graham J . Since 2008, insurers have been required by law to cover mental health—why many still don’t. Atlantic [serial on the Internet]. 2013 Mar 11 [cited 2014 Jun 24 ]. Available from: http://www.theatlantic.com/health/archive/2013/03/since-2008-insurers-have-been-required-by-law-to-cover-mental-health-why-many-still-dont/273562/ Google Scholar
• 39 Internal Revenue Service, Department of the Treasury; Employee Benefits Security Administration, Department of Labor; Centers for Medicare and Medicaid Services, Department of Health and Human Services . Final rules under the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity Act of 2008; technical amendment to external review for multi-state plan program. Final rules . Fed Regist . 2013 ; 78 ( 2 ): 68239 – 96 . Medline, Google Scholar
• 40 Congressional Budget Office . Updated estimates of the effects of the insurance coverage provisions of the Affordable Care Act [Internet]. Washington (DC) : CBO ; 2014 Apr [cited 2014 Jun 24 ]. Appendix B: updated estimates of the insurance coverage provisions of the Affordable Care Act. Available from: http://www.cbo.gov/sites/default/files/cbofiles/attachments/45010-breakout-AppendixB.pdf Google Scholar
• 41 Cummings JR , Wen H , Ko M , Druss BG . Race/ethnicity and geographic access to Medicaid substance use disorder treatment facilities in the United States . JAMA Psychiatry . 2014 ; 71 ( 2 ): 190 – 6 . Crossref, Medline, Google Scholar