Create an Account
OpenAthens Shibboleth
Figure.
Trend in Substance Use Disorder (SUD) Treatment Rate by SUD Parity Status

A, Any change in parity compared with no change. B, All changes in parity ranging from none to full. State-aggregate SUD treatment rates reflect the preparity and postparity periods. We centered the year each state started to implement the law at time 0. The vertical line represents the year during which each state started to implement or extend the law, and it corresponds with the period covered in National Survey of Substance Abuse Treatment Services (N-SSATS) from April 2001 to March 2002 (N-SSATS 2002) for Delaware and Michigan; April 2002 to March 2003 (N-SSATS 2003) for Montana and Rhode Island; April 2003 to March 2004 (N-SSATS 2004) for Maine and New Hampshire; April 2004 to March 2005 (N-SSATS 2005) for Wisconsin and West Virginia; and April 2006 to March 2007 (N-SSATS 2007) for Oregon. Kentucky implemented parity during the gap year between N-SSATS 2000 and 2002, so time 0 consisted of 9 data points instead of 10. For the other states with no change in parity, the treatment rates during N-SSAT 2002, 2003, 2004, 2005, and 2007 were weighted by 2/9, 2/9, 2/9, 2/9, and 1/9, respectively, to match the proportions of the states that implemented parity in a given year. Following the same procedure, we determined times −2, −1, 1, 2, and 3 for the states with any change in parity, and then transferred states with no change in parity to the corresponding time in accord with the parity states. Only 7 states with any change in parity were included for time −2 (no data for Kentucky, Deleware, or Michigan), time 2 (no data for Oregon), and time 3 (no data for Oregon).

Table 1.
Summary of the State Parity Laws for SUD Treatment, 2000-2008
Table 2.
Descriptive Statistics for States Implementing or Extending SUD Parity Laws From 2000 to 2008 vs No Change in SUD Parity Status
Table 3.
Regression Results for the Estimated Effect of SUD Parity Implementation and Other Covariates on the SUD Treatment Rate
1.
Substance Abuse and Mental Health Services Administration. Results From the 2010 National Survey on Drug Use and Health: Summary of National Findings. Rockville, MD: US Dept of Health and Human Services Substance Abuse and Mental Health Services Administration Center for Behaviorial Health Statistics and Quality; 2011. NSDUH Series H-41, HHS Publication (SMA) 11-4658. www.samhsa.gov/data/NSDUH/2k10ResultsRev/NSDUHresultsRev2010.htm. Accessed June 20, 2012.
2.
Winklbaur  B, Jagsch  R, Ebner  N, Thau  K, Fischer  G.  Quality of life in patients receiving opioid maintenance therapy. Eur Addict Res. 2008;14(2):99-105.
PubMedArticle
3.
Roux  P, Carrieri  MP, Villes  V,  et al; MANIF2000 Cohort Study Group.  The impact of methadone or buprenorphine treatment and ongoing injection on highly active antiretroviral therapy (HAART) adherence: evidence from the MANIF2000 Cohort Study. Addiction. 2008;103(11):1828-1836.
PubMedArticle
4.
Mojtabai  R, Zivin  JG.  Effectiveness and cost-effectiveness of four treatment modalities for substance disorders. Health Serv Res. 2003;38(1, pt 1):233-259.Article
5.
Tanner-Smith  EE, Wilson  SJ, Lipsey  MW.  The comparative effectiveness of outpatient treatment for adolescent substance abuse: a meta-analysis. J Subst Abuse Treat. 2013;44(2):145-158.
PubMedArticle
6.
Greenfield  SF, Brooks  AJ, Gordon  SM,  et al.  Substance abuse treatment entry, retention, and outcome in women. Drug Alcohol Depend. 2007;86(1):1-21.
PubMedArticle
7.
Fiellin  DA, Weiss  L, Botsko  M,  et al; BHIVES Collaborative.  Drug treatment outcomes among HIV-infected opioid-dependent patients receiving buprenorphine/naloxone. J Acquir Immune Defic Syndr. 2011;56(suppl 1):S33-S38.
PubMedArticle
8.
Dismuke  CE, French  MT, Salomé  HJ, Foss  MA, Scott  CK, Dennis  ML.  Out of touch or on the money: do the clinical objectives of addiction treatment coincide with economic evaluation results? J Subst Abuse Treat. 2004;27(3):253-263.
PubMedArticle
9.
Sindelar  JL, Jofre-Bonet  M, French  MT, McLellan  AT.  Cost-effectiveness analysis of addiction treatment. Drug Alcohol Depend. 2004;73(1):41-50.
PubMedArticle
10.
Zarkin  GA, Bray  JW, Aldridge  A,  et al.  The effect of alcohol treatment on social costs of alcohol dependence: results from the COMBINE Study. Med Care. 2010;48(5):396-401.
PubMedArticle
11.
McCollister  KE, French  MT.  The relative contribution of outcome domains in the total economic benefit of addiction interventions: a review of first findings. Addiction. 2003;98(12):1647-1659.
PubMedArticle
12.
Hubbard  RL, Craddock  SG, Anderson  J.  Overview of 5-year followup outcomes in the Drug Abuse Treatment Outcome Studies (DATOS). J Subst Abuse Treat. 2003;25(3):125-134.Article
13.
Parran  TV, Adelman  CA, Merkin  B,  et al.  Long-term outcomes of office-based buprenorphine/naloxone maintenance therapy. Drug Alcohol Depend. 2010;106(1):56-60.
PubMedArticle
14.
French  MT, Salomé  HJ, Sindelar  JL, McLellan  AT.  Benefit-cost analysis of addiction treatment: methodological guidelines and empirical application using the DATCAP and ASI. Health Serv Res. 2002;37(2):433-455.
PubMedArticle
15.
Bouchery  EE, Harwood  HJ, Dilonardo  J, Vandivort-Warren  R.  Type of health insurance and the substance abuse treatment gap. J Subst Abuse Treat. 2012;42(3):289-300.
PubMedArticle
16.
Horgan  CM, Merrick  EL.  Financing of substance abuse treatment services. Recent Dev Alcohol. 2001;15:229-252.
17.
Amaro  H.  An expensive policy: the impact of inadequate funding for substance abuse treatment. Am J Public Health. 1999;89(5):657-659.
PubMedArticle
18.
Sturm  R, Sherbourne  CD.  Are barriers to mental health and substance abuse care still rising? J Behav Health Serv Res. 2001;28(1):81-88.
PubMedArticle
19.
Robinson  GC, Whitter  M, Magaña  C. State Mandates for Treatment for Mental Illness and Substance Use Disorders. Rockville, MD: Center for Mental Health Services, Substance Abuse and Mental Health Services Administration; 2006.
20.
Busch  SH.  Implications of the Mental Health Parity and Addiction Equity Act. Am J Psychiatry. 2012;169(1):1-3.
PubMed
21.
Barry  CL, Huskamp  HA.  Moving beyond parity: mental health and addiction care under the ACA. N Engl J Med. 2011;365(11):973-975.
PubMedArticle
22.
Rosenbach ML, Lake T, Young C, et al. Effects of the Vermont Mental Health and Substance Abuse Parity Law. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2003. www.mathematica-mpr.com/PDFs/vtparity.pdf. Accessed June 20, 2012.
23.
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):340-346.
PubMedArticle
24.
Azzone  V, Frank  RG, Normand  SL, Burnam  MA.  Effect of insurance parity on substance abuse treatment. Psychiatr Serv. 2011;62(2):129-134.
PubMedArticle
25.
Dobkin  C, Nicosia  N.  The war on drugs: methamphetamine, public health, and crime. Am Econ Rev. 2009;99(1):324-349.
PubMedArticle
26.
Buck  JA, Teich  JL, Umland  B, Stein  M.  Behavioral health benefits in employer-sponsored health plans, 1997. Health Aff (Millwood). 1999;18(2):67-78.
PubMedArticle
27.
Substance Abuse and Mental Health Services Administration. National Survey of Substance Abuse Treatment Services (N-SSATS), 2009 (ICPSR 28781). Substance Abuse and Mental Health Data Archive. www.icpsr.umich.edu/icpsrweb/SAMHDA/studies/2878. Accessed June 1, 2012.
28.
Hennessy  KD, Goldman  HH.  Full parity: steps toward treatment equity for mental and addictive disorders. Health Aff (Millwood). 2001;20(4):58-67.
PubMedArticle
29.
Cook  BL, Alegría  M.  Racial-ethnic disparities in substance abuse treatment: the role of criminal history and socioeconomic status. Psychiatr Serv. 2011;62(11):1273-1281.
PubMed
30.
Cummings  JR, Wen  H, Druss  BG.  Racial/ethnic differences in treatment for substance use disorders among US adolescents. J Am Acad Child Adolesc Psychiatry. 2011;50(12):1265-1274.Article
31.
Buck  JA.  The looming expansion and transformation of public substance abuse treatment under the Affordable Care Act. Health Aff (Millwood). 2011;30(8):1402-1410.
PubMedArticle
32.
Chirba-Martin  MA, Brennan  TA.  The critical role of ERISA in state health reform. Health Aff (Millwood). 1994;13(2):142-156.Article
33.
Buchmueller  TC, Cooper  PF, Jacobson  M, Zuvekas  SH.  Parity for whom? Health Aff (Millwood). 2007;26(4):w483-w487.
PubMedArticle
34.
Wooldridge  JM. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press; 2001.
35.
Sturm  R, Pacula  RL.  State mental health parity laws. Health Aff (Millwood). 1999;18(5):182-192.
PubMedArticle
36.
Bertrand M, Duflo E, Mullainathan S. How much should we trust differences-in-differences estimates? Cambridge, MA: National Bureau of Economic Research; March 2002. NBER Working Paper 8841. http://ideas.repec.org/p/nbr/nberwo/8841.html. Accessed August 5, 2012.
37.
McConnell  KJ, Gast  SHN, Ridgely  MS,  et al.  Behavioral health insurance parity: does Oregon’s experience presage the national experience with the Mental Health Parity and Addiction Equity Act? Am J Psychiatry. 2012;169(1):31-38.
PubMed
38.
Goldman  W, McCulloch  J, Sturm  R.  Costs and use of mental health services before and after managed care. Health Aff (Millwood). 1998;17(2):40-52.
PubMedArticle
39.
Ma  CA, McGuire  TG.  Costs and incentives in a behavioral health carve-out. Health Aff (Millwood). 1998;17(2):53-69.
PubMedArticle
Views 3,094
• Cite This

### Citation

Wen H, Cummings JR, Hockenberry JM, Gaydos LM, Druss BG. State Parity Laws and Access to Treatment for Substance Use Disorder in the United StatesImplications for Federal Parity Legislation. JAMA Psychiatry. 2013;70(12):1355-1362. doi:10.1001/jamapsychiatry.2013.2169

Original Investigation
December 2013

# State Parity Laws and Access to Treatment for Substance Use Disorder in the United StatesImplications for Federal Parity Legislation

Author Affiliations
• 1Department of Health Policy and Management, Rollins School of Public Health, Emory University, Atlanta, Georgia
• 2currently with Center for Comprehensive Access and Delivery Research and Evaluation Center, Iowa City Veterans Affairs Health Care System, Iowa City, Iowa
JAMA Psychiatry. 2013;70(12):1355-1362. doi:10.1001/jamapsychiatry.2013.2169
Abstract

Importance  The passage of the 2008 Mental Health Parity and Addiction Equity Act and the 2010 Affordable Care Act incorporated parity for substance use disorder (SUD) treatment into federal legislation. However, prior research provides us with scant evidence as to whether federal parity legislation will hold the potential for improving access to SUD treatment.

Objective  To examine the effect of state-level SUD parity laws on state-aggregate SUD treatment rates and to shed light on the impact of the recent federal SUD parity legislation.

Design, Setting, and Participants  We conducted a quasi-experimental study using a 2-way (state and year) fixed-effect method. We included all known specialty SUD treatment facilities in the United States and examined treatment rates from October 1, 2000, through March 31, 2008. Our main source of data was the National Survey of Substance Abuse Treatment Services, which provides facility-level information on specialty SUD treatment.

Interventions  State-level SUD parity laws during the study period.

Main Outcomes and Measures  State-aggregate SUD treatment rates in (1) all specialty SUD treatment facilities and (2) specialty SUD treatment facilities accepting private insurance.

Results  The implementation of any SUD parity law increased the treatment rate by 9% (P < .001) in all specialty SUD treatment facilities and by 15% (P = .02) in facilities accepting private insurance. Full parity and parity only if SUD coverage is offered increased the SUD treatment rate by 13% (P = .02) and 8% (P = .04), respectively, in all facilities and by 21% (P = .03) and 10% (P = .04), respectively, in facilities accepting private insurance.

Conclusions and Relevance  We found a positive effect of the implementation of state SUD parity legislation on access to specialty SUD treatment. Furthermore, the positive association is more pronounced in states with more comprehensive parity laws. Our findings suggest that federal parity legislation holds the potential to improve access to SUD treatment.

An estimated 23 million Americans had a substance use disorder (SUD) in 2010, including abuse of or dependence on alcohol and/or illicit drugs.1 A growing body of literature has demonstrated the efficacy and cost-effectiveness of treatment for SUD. Specialty SUD treatment services such as outpatient psychosocial therapy and opioid maintenance therapy have proved to be effective in improving health,28 reducing crime,46,911 increasing employment,8,9,1214 and producing a wide range of social benefits.5,6,9,1214 Nonetheless, only 17% of those individuals who needed SUD treatment received any treatment for their condition, and only 11% (2.6 million) received treatment in a specialty setting.1

Financial barriers in general and limited insurance coverage for SUD in particular pose a major barrier to access to specialty SUD treatment among those individuals perceiving a need for treatment.1,15 Ever since the inception of third-party payment for SUD treatment, coverage for SUD treatment has been more restrictive than that for medical/surgical treatment in terms of cost sharing and treatment limitations.1618 To address these discriminatory restrictions, more than one-half of the states in the United States have enacted SUD parity laws during the past 2 decades requiring employment-related group health plans to provide coverage for SUD treatment equal to that for comparable medical/surgical treatment.19

More recently, the passage of the 2008 Mental Health Parity and Addiction Equity Act (MHPAEA) incorporated SUD parity into federal legislation for the first time.20 However, the MHPAEA mandates parity only for employment-related and self-funded group health plans and only for existing SUD coverage offered by those plans (ie, parity-if-offered). Subsequently, provisions of the 2010 Affordable Care Act (ACA) extended SUD parity to Medicaid-managed care plans, Medicaid benchmark and benchmark-equivalent plans, and state health insurance exchange plans.21 Furthermore, the ACA requires that coverage for SUD treatment, as an essential health benefit, must be offered and must be offered on par with that for comparable medical/surgical treatment (ie, full parity).

Nonetheless, prior research provides us with scant evidence about the likely effect of federal parity legislation on access to SUD treatment. Two studies examined SUD parity laws in the private insurance market of a particular state (ie, Vermont and Oregon),22,23 and a third study evaluated SUD parity implementation in the Federal Employees Health Benefits program24; none of these studies found a significant improvement in access to SUD treatment attributable to the implementation of SUD parity. However, findings from these studies may have limited generalizability to the anticipated effect of the recent federal SUD parity legislation. First, the study examining Vermont’s 1998 parity law22 did not include a comparison group to control for the downward secular trend in access to SUD treatment nationwide. In addition, the study examining Oregon’s 2007 parity law23 captured only a policy change from partial parity (implemented in 2000) to full parity, which might be confounded by Oregon’s simultaneous reform of methamphetamine regulation (effective in July 2006) that dramatically curbed the underlying prevalence rate.25 Finally, the study evaluating parity of the Federal Employees Health Benefits focused on a study population with a unique risk profile (eg, less likely to use and abuse or to depend on substance) and financial capacity (eg, less likely to have financial barriers to treatment) that may limit the generalizability of the results to broader populations with private insurance.26

The present study advances the existing literature by analyzing all state-level SUD parity laws in the private insurance market implemented from October 1, 2000, through March 31, 2008, and applying a rigorous quasi-experimental design to the variations among those state parity laws in the timing of the implementation and the comprehensiveness of the mandate. We hypothesized that (1) the implementation of SUD parity legislation increased the SUD treatment rate at the state level; (2) the increase in the treatment rate was more pronounced in states with more comprehensive SUD parity laws; and (3) the increase in the SUD treatment rate associated with the implementation of SUD parity laws was concentrated in facilities accepting private insurance.

Methods
Data Sources

The main source of our data is the National Survey of Substance Abuse Treatment Services (N-SSATS),27 which provides facility-level information on specialty SUD treatment from 2000 through 2008. (In 2002, the reference date for the annual survey was changed from September to March to enhance the response rate, leaving a gap period from September 2000 to March 2001 with no data collected.) The N-SSATS facility data cover all known specialty SUD treatment facilities, allowing for a nearly complete enumeration of specialty SUD treatment services in the United States. A specialty SUD treatment facility, according to N-SSATS, is defined as a hospital, a residential SUD facility, an outpatient SUD treatment facility, a mental health facility with an SUD treatment program, or other facility with an SUD treatment program providing the following treatment services: (1) outpatient, inpatient, or residential/rehabilitation SUD treatment; (2) detoxification treatment; (3) opioid treatment programs such as maintenance therapy with methadone and levo-α-acetylmethadol; and (4) halfway-house services that include SUD treatment. Throughout the study period, response rates ranged from 92% to 95%.27 We merged the N-SSATS data with select state-level measures from nationally representative data sets to provide supplementary information on important state socioeconomic characteristics and policy environment (discussed below). The Emory University Institutional Review Board determined that this study did not require institutional review board oversight because all data were collected from publically available sources and de-identified, and the authors did not have access to any protected health information.

Analytic Sample

We combined the N-SSATS data sets from 2000 to 2008 and converted the facility-level data to the state level to create an analytic panel of 392 state-year observations across the 49 states and 8 years. Virginia was excluded from the analysis because it was the only state that moved away from parity when full parity was repealed and regressed to partial parity in 2004.

Measurement of Variables
Dependent Variable

All surveyed facilities were requested to report the total SUD treatment counts in the most recent 12 months before the survey. The N-SSATS method specified that the treatment count should only include the initial entry of a client into treatment; subsequent visits to the same service or transfer to a different service within a single continuous course of treatment were excluded. The missing-item rate for treatment count was approximately 7% during the study period.

The treatment counts in all specialty facilities were aggregated to each state s in each year t to determine the state-aggregate annual number of SUD treatment entries. We also aggregated the treatment counts only for facilities that accept private insurance. Both measures of the state-aggregate annual treatment entries were then weighted by the state population size to generate the 2 dependent variables assessing the following: (1) the treatment rate among all facilities (treatment ratest) as the number of SUD treatment entries into all specialty SUD treatment facilities per 100 state residents in each state s for each year t; and (2) the treatment rate for facilities accepting private insurance (treatment rate-PIst) as the number of SUD treatment entries into specialty SUD treatment facilities that accept private insurance per 100 state residents in each state s for each year t.

Primary Independent Variables

In a broad sense, SUD parity refers to a policy mandating insurance coverage for SUD treatment to be “no more restrictive” than coverage for comparable medical/surgical treatment, with respect to cost sharing (eg, deductibles, copayments, coinsurance, and out-of-pocket expenses), treatment limitations (eg, annual or lifetime limits on number of visits or hospital days), or both.28 The first independent variable of interest is a dichotomous indicator for the implementation of any parity law in a given state s during a given year t (parityst). The implementation indicator was assigned a value of 1 for each full year subsequent to the time when a state first implemented its SUD parity law and a value of 0 for the preimplementation periods and for states without any SUD parity law.

We also created a categorical measure to distinguish among the following different levels of comprehensiveness in the implementation of parity:

(1) Full parityst requires SUD coverage to be offered and offered on par with the comparable medical/surgical coverage in all aspects of cost sharing and treatment limitations;

(2) Partial parityst, though requiring that SUD coverage be offered, allows for discrepancies between SUD coverage and comparable medical/surgical coverage in some aspects of cost sharing and treatment limitation;

(3) Parity-if-offeredst does not require SUD coverage to be offered, but if offered, it should be on par with the comparable medical/surgical coverage in all aspects of cost sharing and treatment limitations.

To assess the implementation and the comprehensiveness of the state SUD parity laws, we reviewed the relevant information provided by the Substance Abuse and Mental Health Services Administration,19 the National Conference of State Legislatures, and other advocacy organizations. We also referred to the original state statutes to detect the subtlety in statutory language and to reconcile the inconsistencies among various sources. Table 1 presents detailed information on state SUD parity laws during the study period.

Covariates

To account for the state-year heterogeneity, we included key time-varying sociodemographic characteristics and policy environment factors that have been extensively documented to influence access to SUD treatment.15,29,30 Our covariates constituted the percentage of state population who are (1) black or African American, (2) Hispanic or Latino, (3) living in poverty (≤100% at or below the federal poverty line), (4) classified with SUD (ie, meets the DSM-IV diagnostic criteria for alcohol abuse/dependence and/or illicit drug abuse/dependence), and (5) eligible for Medicaid. We also included the per capita amount of the Substance Abuse Prevention and Treatment Block Grant (SAPTBG) allocated to the state as a proxy for system capacity.31 The SAPTBG represents a significant federal contribution to the state budgets for substance abuse prevention and treatment systems and accounts for approximately 40% of public funds expended by states for SUD treatment. In 2001, 16 states reported that more than half of their total funding for SUD treatment programs came from the SAPTBG.

In addition to the sociodemographic and policy covariates, we adjusted for the target population and exemption conditions that are commonly included in state SUD parity legislation. Most parity laws apply only to employment-related group health plans, leaving the individual (non–employment-based) health insurance market unregulated. Moreover, the federal preemption by the Employee Retirement Income Security Act of 197432 does not allow state legislatures to impose health insurance regulations on self-insured business (usually large employers). Some states also exempt employers with fewer than 50 or fewer than 20 employees, further limiting the reach of SUD parity.33 When we considered the availability of the consistent data across the study states and years, we controlled for the percentage of the state population (1) covered by employer-sponsored health insurance, (2) covered by individually purchased health insurance, (3) with large employers (ie, >500 employees), and (4) with small employers (ie, <20 employees).

Statistical Analysis

We analyzed the effect of state SUD parity laws on state-aggregate SUD treatment rates, using 2-way (state and year) fixed-effect modeling to account for unobserved or unmeasured factors in the treatment rates that are systematically correlated with the parity laws. The 2-way fixed-effect approach can be viewed as an extension of the difference-in-difference framework to fit multiple-unit and multiple-time models that go beyond the traditional 2 groups (intervention vs comparison) and 2 periods (before and after).34 By distinguishing the real impact of parity legislation from the confounding factors of the state heterogeneity35 and the national secular trend, we are able to obtain unbiased estimates of the effect of state SUD parity laws.

We estimated 4 models. Model 1 estimated the SUD treatment rate among all specialty SUD treatment facilities at state s in year t (treatment ratest) as a function of the dichotomous indicator of SUD parity implementation (parityst), the state fixed effect (υs), the year fixed effect (τt), the state sociodemographic and policy covariates (covariate vectorst), and an idiosyncratic error term (εst). Model 2 replaced the dichotomous indicator of any SUD parity implementation (parityst) with the categorical variable of the comprehensiveness in parity mandate (full parityst, partial parityst, and parity-if-offeredst). The dependent variable of both models, the SUD treatment rate among all specialty SUD treatment facilities (treatment ratest), was measured based on the entire population instead of those targeted by state parity. The estimated effect of parity legislation, in this sense, would be diluted over a mixture of target (ie, those groups with private insurance plans affected by parity) and nontarget groups (ie, those groups with no insurance, with public insurance, or with private insurance plans not affected by parity). To refine our crude estimates, we also limited the treatment rate measure to facilities accepting private insurance (treatment rate-PIst) and reestimated the 2 models described above.

All estimated standard errors were clustered at the state level to correct for the serial correlation that otherwise leads to false rejections of the null hypothesis.36 Analysis was performed using commercially available statistical software (STATA, version 12.0; StataCorp).

Results

The Figure shows an upward trend in the SUD treatment rate in parallel with the implementation of SUD parity legislation. Among the 10 states that first implemented SUD parity or extended their parity laws to a higher level of comprehensiveness from 2000 through 2008, the mean SUD treatment rate rose from 1.38 percentage points (per 100 population) during the year immediately before the parity implementation to 1.53 percentage points in the year immediately after implementation. The preparity and postparity change in the SUD treatment rate was equivalent to an 11% increase ([1.53 − 1.38] ÷ {[1.53 + 1.38] ÷ 2}). Among states that did not change their SUD parity status, the mean SUD treatment rate fell from 1.44 to 1.38 percentage points during the same period, which corresponds to an decrease of 4% ([1.38 − 1.44] ÷ {[1.38 + 1.44] ÷ 2}). This observational trend comparison demonstrated a positive association between SUD parity and treatment rate.

Table 2 summarizes the descriptive statistics for the following 3 groups of states: (1) the 10 parity states that first implemented parity laws or extended their laws from 2000 through 2008; (2) the 23 states that do not have SUD parity; and (3) the other 16 states that first implemented parity laws before 2000 and did not change their laws during the study period. We combined groups 2 and 3 as the control group representing the states without changes in parity laws during the study period. The 2-sample t tests of mean differences between the 10 parity states with changes in their parity laws and the remaining states without changes indicated that the parity states had a significantly higher rate of SUD treatment in all specialty SUD treatment facilities (P = .03) and in facilities accepting private insurance (P < .001).

Table 3 reports the regression results for the estimated effect of SUD parity implementation on the SUD treatment rate. The implementation of any SUD parity law significantly increased the treatment rate in all specialty SUD treatment facilities (model 1.1, marginal effect [ME] = 0.13 percentage points [95% CI, 0.04-0.23]) and in facilities accepting private insurance (model 2.1, 0.16 [0.03-0.30]). To place the magnitude of effect into context, we translate the estimated ME (ie, change in percentage points per 100 state residents) into the percentage of change in the SUD treatment rate. Given that the mean SUD treatment rate was 1.40 percentage points in all specialty SUD treatment facilities and 1.10 percentage points in facilities accepting private insurance, changes of 0.13 and 0.16 percentage points, respectively, can be translated into a 9% increase in the overall SUD treatment rate (ie, 9% = 0.13 ÷ 1.40), and a 15% increase in the SUD treatment rate for facilities accepting private insurance (ie, 15% = 0.16 ÷ 1.10).

When considering the comprehensiveness of the parity legislation (Table 3), full parity and parity-if-offered increased the SUD treatment rate by 13% (model 1.2, ME = 0.18 percentage points [95% CI, 0.03-0.33]) and 8% (model 1.2, 0.12 [0.00-0.23]), respectively, in all facilities and by 21% (model 2.2, 0.23 [0.03-0.43]) and 10% (model 2.2, 0.11 [0.00-0.22]), respectively, in those accepting private insurance. The influence of partial parity on the treatment rate was not statistically significant across models.

Discussion

Our findings indicate that the implementation of state SUD parity legislation results in a significant improvement in access to specialty SUD treatment. The implementation of any SUD parity law increased the treatment rate by 9% in all specialty SUD treatment facilities and by 15% in facilities accepting private insurance. Our study contributes to the existing literature by using state-level panel data on a nearly complete enumeration of all treatment counts in specialty SUD treatment facilities, harnessing all legislative changes in state-level SUD parity laws during the study period, and tailoring a rigorous quasi-experimental design to this series of state experiments.

Our study also advances the literature by documenting the extent to which the comprehensiveness of SUD parity matters. The implementation of full parity laws led to the largest increases in SUD treatment rate (a 13% increase), followed by parity-if-offered laws (an 8% increase). The effect of partial parity, on the other hand, was not statistically significant (P = .12).

When considering the implications of our findings for the anticipated impact of recent federal SUD parity legislation, the MHPAEA (ie, parity-if-offered) can be expected to have a modest effect on access to SUD treatment. The MHPAEA not only regulates quantitative limits (eg, annual or lifetime limits on the number of visits or hospital days) addressed by previous state-level parity laws but also mandates parity for a wider range of nonquantitative restrictions such as medical necessity, prior authorization, or utilization review.23,3739 Given the dominance of these managed care mechanisms in the SUD service arrangements of private health plans, the inclusion of the nonquantitative managed care restrictions into the MHPAEA may enable this legislation to yield larger effects on the SUD treatment rate than we estimated for the state-level parity-if-offered laws.

Under the ACA, the full-parity provision, coupled with insurance expansion, is likely to further improve the access to SUD treatment beyond the impact of state-level full-parity laws. The ACA will expand health insurance to approximately 50 million uninsured persons; SUD coverage gained by the newly insured persons through Medicaid benchmark or benchmark-equivalent plans or state health insurance exchange plans will be subject to full parity.21 In our analysis of the state parity regulations in the employment-related group insurance market, the increases associated with full SUD parity were confined to facilities accepting private insurance. By expanding the scope of parity to public insurance programs, the ACA will reach a much larger population and may lead to an unparalleled growth of the SUD treatment rate in the public and private sectors.

The estimated growth in SUD treatment rate will only be possible if the capacity of the SUD treatment system suffices to absorb new entrants into the system. At present, most SUD treatment is provided in the specialty treatment sector, and researchers have already raised concerns that SUD specialty treatment programs may face challenges in meeting potential needs.31 The Prevention and Public Health Fund created under the ACA offers grant support to develop more comprehensive SUD screening, brief intervention, referral, and treatment programs, which will enhance the capacity of primary care sites to provide SUD care. Enhanced funding for federally qualified health care centers and Medicaid health home initiatives may also help to fill the capacity gap. Nonetheless, as the MHPAEA and ACA unfold, tracking the effect of both laws on SUD treatment to ensure that they are able to fulfill their promise in improving access to SUD treatmentwill be critical.

The conclusions of this study should be interpreted in light of the following limitations. First, we cannot identify individuals’ insurance coverage and their employment status in the facility-level N-SSATS data or find more detailed facility-level information on the percentage of treatment entries/clients who were covered by the health insurance plans subject to parity. Thus, the dependent variable, the state-aggregate SUD treatment rate, was measured based on the entire population instead of the population targeted by state parity laws. We refined our analysis by restricting the measurement of the treatment rate to facilities accepting private insurance, which yielded a larger point estimate of the parity effect. We also conducted sensitivity analyses for facilities not accepting private insurance (Supplement [eTables 1-6]) and found no difference in SUD treatment rates attributable to parity. Considered together, these additional analyses suggest that the effect of SUD parity on the treatment rate is primarily driven by the increased treatment rate among the target population. Second, N-SSATS did not ask facilities to report treatment counts for alcohol and illicit drug use separately; thus, we were only able to assess the effect of parity on combined SUD treatment rates, despite their distinct legal status, patterns of treatment, and consequently individuals’ policy sensitivity and price elasticity. Third, as with any observational study, we cannot definitively establish causality between the implementation of SUD parity laws and access to SUD treatment. However, the rigorous methods and robust results strongly suggest that parity improved access.

Despite these limitations, our study provides useful insight into the potential effect of the implementation and the comprehensiveness of SUD parity on access to SUD treatment and, in broad terms, the potential of financial incentives and policy leverage to influence treatment-seeking behavior. We found that the implementation of state SUD parity laws significantly increased the SUD treatment rate and that the increase was more pronounced in states implementing more comprehensive laws. These findings suggest that the MHPAEA of 2008 and the ACA of 2010 hold the potential to improve access to SUD treatment.

Article Information

Submitted for Publication: November 12, 2013; final revision received February 12, 2013; accepted March 25, 2013.

Corresponding Author: Hefei Wen, BA, Department of Health Policy and Management, Rollins School of Public Health, Emory University, 1518 Clifton Rd NE, Room 623, Atlanta, GA 30322 (hwen2@emory.edu).

Published Online: October 23, 2013. doi:10.1001/jamapsychiatry.2013.2169.

Author Contributions: Ms Wen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Wen, Cummings, Gaydos, Druss.

Acquisition of data: Wen.

Analysis and interpretation of data: Wen, Cummings, Hockenberry, Druss.

Drafting of the manuscript: Wen.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Wen, Hockenberry.

Obtained funding: Cummings, Druss.

Administrative, technical, or material support: Cummings and Gaydos.

Study supervision: Cummings, Gaydos.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported in part by grants 5K24MH07586705 and 1K01MH09582301 from the National Institute of Mental Health (Drs Cummings and Druss).

Role of the Sponsor: The National Institute of Mental Health had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

References
1.
Substance Abuse and Mental Health Services Administration. Results From the 2010 National Survey on Drug Use and Health: Summary of National Findings. Rockville, MD: US Dept of Health and Human Services Substance Abuse and Mental Health Services Administration Center for Behaviorial Health Statistics and Quality; 2011. NSDUH Series H-41, HHS Publication (SMA) 11-4658. www.samhsa.gov/data/NSDUH/2k10ResultsRev/NSDUHresultsRev2010.htm. Accessed June 20, 2012.
2.
Winklbaur  B, Jagsch  R, Ebner  N, Thau  K, Fischer  G.  Quality of life in patients receiving opioid maintenance therapy. Eur Addict Res. 2008;14(2):99-105.
PubMedArticle
3.
Roux  P, Carrieri  MP, Villes  V,  et al; MANIF2000 Cohort Study Group.  The impact of methadone or buprenorphine treatment and ongoing injection on highly active antiretroviral therapy (HAART) adherence: evidence from the MANIF2000 Cohort Study. Addiction. 2008;103(11):1828-1836.
PubMedArticle
4.
Mojtabai  R, Zivin  JG.  Effectiveness and cost-effectiveness of four treatment modalities for substance disorders. Health Serv Res. 2003;38(1, pt 1):233-259.Article
5.
Tanner-Smith  EE, Wilson  SJ, Lipsey  MW.  The comparative effectiveness of outpatient treatment for adolescent substance abuse: a meta-analysis. J Subst Abuse Treat. 2013;44(2):145-158.
PubMedArticle
6.
Greenfield  SF, Brooks  AJ, Gordon  SM,  et al.  Substance abuse treatment entry, retention, and outcome in women. Drug Alcohol Depend. 2007;86(1):1-21.
PubMedArticle
7.
Fiellin  DA, Weiss  L, Botsko  M,  et al; BHIVES Collaborative.  Drug treatment outcomes among HIV-infected opioid-dependent patients receiving buprenorphine/naloxone. J Acquir Immune Defic Syndr. 2011;56(suppl 1):S33-S38.
PubMedArticle
8.
Dismuke  CE, French  MT, Salomé  HJ, Foss  MA, Scott  CK, Dennis  ML.  Out of touch or on the money: do the clinical objectives of addiction treatment coincide with economic evaluation results? J Subst Abuse Treat. 2004;27(3):253-263.
PubMedArticle
9.
Sindelar  JL, Jofre-Bonet  M, French  MT, McLellan  AT.  Cost-effectiveness analysis of addiction treatment. Drug Alcohol Depend. 2004;73(1):41-50.
PubMedArticle
10.
Zarkin  GA, Bray  JW, Aldridge  A,  et al.  The effect of alcohol treatment on social costs of alcohol dependence: results from the COMBINE Study. Med Care. 2010;48(5):396-401.
PubMedArticle
11.
McCollister  KE, French  MT.  The relative contribution of outcome domains in the total economic benefit of addiction interventions: a review of first findings. Addiction. 2003;98(12):1647-1659.
PubMedArticle
12.
Hubbard  RL, Craddock  SG, Anderson  J.  Overview of 5-year followup outcomes in the Drug Abuse Treatment Outcome Studies (DATOS). J Subst Abuse Treat. 2003;25(3):125-134.Article
13.
Parran  TV, Adelman  CA, Merkin  B,  et al.  Long-term outcomes of office-based buprenorphine/naloxone maintenance therapy. Drug Alcohol Depend. 2010;106(1):56-60.
PubMedArticle
14.
French  MT, Salomé  HJ, Sindelar  JL, McLellan  AT.  Benefit-cost analysis of addiction treatment: methodological guidelines and empirical application using the DATCAP and ASI. Health Serv Res. 2002;37(2):433-455.
PubMedArticle
15.
Bouchery  EE, Harwood  HJ, Dilonardo  J, Vandivort-Warren  R.  Type of health insurance and the substance abuse treatment gap. J Subst Abuse Treat. 2012;42(3):289-300.
PubMedArticle
16.
Horgan  CM, Merrick  EL.  Financing of substance abuse treatment services. Recent Dev Alcohol. 2001;15:229-252.
17.
Amaro  H.  An expensive policy: the impact of inadequate funding for substance abuse treatment. Am J Public Health. 1999;89(5):657-659.
PubMedArticle
18.
Sturm  R, Sherbourne  CD.  Are barriers to mental health and substance abuse care still rising? J Behav Health Serv Res. 2001;28(1):81-88.
PubMedArticle
19.
Robinson  GC, Whitter  M, Magaña  C. State Mandates for Treatment for Mental Illness and Substance Use Disorders. Rockville, MD: Center for Mental Health Services, Substance Abuse and Mental Health Services Administration; 2006.
20.
Busch  SH.  Implications of the Mental Health Parity and Addiction Equity Act. Am J Psychiatry. 2012;169(1):1-3.
PubMed
21.
Barry  CL, Huskamp  HA.  Moving beyond parity: mental health and addiction care under the ACA. N Engl J Med. 2011;365(11):973-975.
PubMedArticle
22.
Rosenbach ML, Lake T, Young C, et al. Effects of the Vermont Mental Health and Substance Abuse Parity Law. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2003. www.mathematica-mpr.com/PDFs/vtparity.pdf. Accessed June 20, 2012.
23.
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):340-346.
PubMedArticle
24.
Azzone  V, Frank  RG, Normand  SL, Burnam  MA.  Effect of insurance parity on substance abuse treatment. Psychiatr Serv. 2011;62(2):129-134.
PubMedArticle
25.
Dobkin  C, Nicosia  N.  The war on drugs: methamphetamine, public health, and crime. Am Econ Rev. 2009;99(1):324-349.
PubMedArticle
26.
Buck  JA, Teich  JL, Umland  B, Stein  M.  Behavioral health benefits in employer-sponsored health plans, 1997. Health Aff (Millwood). 1999;18(2):67-78.
PubMedArticle
27.
Substance Abuse and Mental Health Services Administration. National Survey of Substance Abuse Treatment Services (N-SSATS), 2009 (ICPSR 28781). Substance Abuse and Mental Health Data Archive. www.icpsr.umich.edu/icpsrweb/SAMHDA/studies/2878. Accessed June 1, 2012.
28.
Hennessy  KD, Goldman  HH.  Full parity: steps toward treatment equity for mental and addictive disorders. Health Aff (Millwood). 2001;20(4):58-67.
PubMedArticle
29.
Cook  BL, Alegría  M.  Racial-ethnic disparities in substance abuse treatment: the role of criminal history and socioeconomic status. Psychiatr Serv. 2011;62(11):1273-1281.
PubMed
30.
Cummings  JR, Wen  H, Druss  BG.  Racial/ethnic differences in treatment for substance use disorders among US adolescents. J Am Acad Child Adolesc Psychiatry. 2011;50(12):1265-1274.Article
31.
Buck  JA.  The looming expansion and transformation of public substance abuse treatment under the Affordable Care Act. Health Aff (Millwood). 2011;30(8):1402-1410.
PubMedArticle
32.
Chirba-Martin  MA, Brennan  TA.  The critical role of ERISA in state health reform. Health Aff (Millwood). 1994;13(2):142-156.Article
33.
Buchmueller  TC, Cooper  PF, Jacobson  M, Zuvekas  SH.  Parity for whom? Health Aff (Millwood). 2007;26(4):w483-w487.
PubMedArticle
34.
Wooldridge  JM. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press; 2001.
35.
Sturm  R, Pacula  RL.  State mental health parity laws. Health Aff (Millwood). 1999;18(5):182-192.
PubMedArticle
36.
Bertrand M, Duflo E, Mullainathan S. How much should we trust differences-in-differences estimates? Cambridge, MA: National Bureau of Economic Research; March 2002. NBER Working Paper 8841. http://ideas.repec.org/p/nbr/nberwo/8841.html. Accessed August 5, 2012.
37.
McConnell  KJ, Gast  SHN, Ridgely  MS,  et al.  Behavioral health insurance parity: does Oregon’s experience presage the national experience with the Mental Health Parity and Addiction Equity Act? Am J Psychiatry. 2012;169(1):31-38.
PubMed
38.
Goldman  W, McCulloch  J, Sturm  R.  Costs and use of mental health services before and after managed care. Health Aff (Millwood). 1998;17(2):40-52.
PubMedArticle
39.
Ma  CA, McGuire  TG.  Costs and incentives in a behavioral health carve-out. Health Aff (Millwood). 1998;17(2):53-69.
PubMedArticle