A. Purpose of This Report
Expansion of insurance coverage for substance use disorders (SUDs) under the Affordable Care Act (ACA) and Mental Health Parity and Addiction Equity Act (MHPAEA) may offer opportunity to improve access to care and reduce the societal costs related to SUDs. Projections from the Congressional Budget Office (CBO 2014) indicate the ACA will result in substantial increases in the number of individuals with insurance coverage obtained through the marketplaces or Medicaid through 2020. In addition, the U.S. Department of Health and Human Services (HHS) Substance Abuse and Mental Health Services Administration (SAMHSA 2013d) projects many of these individuals will have SUD treatment needs. CBO forecasts that 13 million people will obtain coverage though marketplaces in 2015, 24 million in 2016 when more substantial tax penalties are implemented, and 25 million in 2017-2024. SAMHSA estimates that approximately 15 percent of uninsured adults who would likely be eligible for subsidized coverage through the marketplaces meet the criteria for a SUD. Additionally, as a result of the ACA, CBO projects 11 million people will obtain Medicaid or Children's Health Insurance Program (CHIP) in 2015, 12 million in 2016 and 2017, and 13 million in 2018-2024. SAMHSA forecasts that roughly 14 percent of uninsured adults who meet eligibility for expanded Medicaid coverage meet the criteria an SUD. Federal officials are concerned that the existing SUD treatment system will not be able to meet the increased demand for treatment, because the existing workforce is insufficient.
To address this concern, in September 2014, the HHS Office of the Assistant Secretary for Planning and Evaluation (ASPE) contracted with Mathematica Policy Research to conduct this project to assess: (1) current demand for SUD treatment; (2) how demand will change as more people obtain SUD treatment coverage; (3) the current state of provider capacity in the SUD treatment field; and (4) the degree to which SUD treatment providers are prepared to be more integrated with the broader health care system. This report summarizes the findings from the first phase of this study which encompassed interviews with experts from two provider credentialing organizations and a national provider organization representative, as well as a review of the available literature and data on SUD prevalence, treatment, and workforce capacity. The second phase of this study entails collection and analysis of survey data on the size and characteristics of the workforce. We expect to complete the second phase of this study in February 2018.
This report summarizes the findings from interviews conducted with experts and a review of the available literature and data analysis on SUD treatment demand and workforce capacity. We discuss the specific methods for these tasks here.
1. Review of Available Literature and Data Analysis
The first step in the literature review was to identify search criteria to address the four topic areas identified for this study:
Current demand for SUD treatment services.
Trends and policies impacting demand over the next decade.
Previous efforts to estimate the size and composition of the SUD workforce.
Recruiting and developing the workforce.
We defined a set of key words addressing these topic areas. Then, we conducted a search of articles indexed in CINAHL, PsycINFO, Scopus, and PubMed from 2005 through November 2014. This search generated a list of potential sources. We also conducted a Google search to identify important studies in the gray literature. We reviewed the publication information and abstracts for potential sources for relevance and strength of analysis. Then, we obtained and reviewed the studies most likely to provide evidence related to the four topic areas of interest. We drew out relevant information from each study and discussed the findings under the appropriate topic areas in this report.
2. Review of Available Data Sources
In addition to reviewing the existing literature, we reviewed the data sources available to support analysis of supply and demand. We discuss the sources identified and their strengths and limitations here.
a. Data Sources for Demand
There is no single data source available to comprehensively estimate SUD treatment demand (Table I.1). The National Survey of Drug Use and Health (NSDUH) isan annual survey of the United States, civilian, non-institutionalized population age 12 and above.4 As a population survey it provides the most comprehensive information about the number of individuals who accessed any service in the past year and the characteristics of those individuals. However the NSDUH provides limited information on the type of services used and no information on the intensity of services used. The National Survey of Substance Abuse Treatment Services (N-SSATS) is a specialty treatment facility survey.5 N-SSATS provides point-in-time estimates of clients in care at specialty facilities by type of service received. Similar to the NSDUH the N-SSATS provides no information on the intensity of services provided to individual users of a given service type. N-SSATS also provides limited information on client characteristics. A strength of the N-SSATS is that it surveys the universe of SUD treatment facilities in the United States and thus provides sufficient data for state-level analysis. Overall we estimate N-SSATS represents more than 72 percent of SUD treatment spending nationally (Table I.1).
Information on services provided in a primary care setting or by independent practitioners is available through National Ambulatory Medical Care Survey (NAMCS) and Medical Expenditure Panel Survey (MEPS). However, in contrast to the NSDUH and N-SSATS these surveys were not specifically designed to capture SUD treatment trends. Therefore, SUD treatment services may be underreported and the sample of SUD treatment services available in the annual samples is limited. The NAMCS data is limited to physician and clinic services. The MEPS addresses physician and clinic services as well as services provided by counselors, nurse practitioners and physician assistants.
b. Data Sources for Supply
We were not able to identify a comprehensive source for estimating the supply of SUD treatment providers (Table I.2). SUD treatment is provided in a variety of treatment settings using the skills of professionals in multiple fields. No single available data source provides a comprehensive enumeration of all SUD treatment providers. The Standard Occupational Classification (SOC) system that is used by the Bureau of Labor Statistics (BLS) to categorize workers for the purpose of collecting and disseminating data on employment and wages does not capture SUD treatment providers as a category separate from other types of providers. For example, physicians and social workers working in the SUD treatment field might be categorized as psychiatrist and mental health and substance abuse social workers, respectively.
Data on the workforce collected through the N-SSATS 2016 survey will be the primary source of estimates of supply for this study. The workforce represented in the N-SSATS survey will include all workers in specialty SUD treatment facilities nationally which represent about three quarters of SUD treatment expenditures. Data from the N-SSATS workforce questions can provide estimates of the workforce by geographic area.
|TABLE I.1. Summary of Available Sources of Demand Data
('x' indicates the service type is represented in the indicated data source)
|Provider Type||% of SUD Treatment Spending 20091||NAMCS||NHAMCS||HCUP-NIS||N-SSATS||MEPS||TEDS||NSDUH|
|Outpatient and residential treatment facilities||46||X||Limited2||Limited3||X|
|Hospital outpatient-- specialty4||9||X||X||X||Limited3||X|
|Hospital residential-- specialty4||3||X||Limited2||Limited3||X|
|Hospital inpatient-- non- specialty4||3||X||X||X|
|Hospital outpatient-- non- specialty4||5||X||X||X|
|Hospital residential-- non- specialty4||<1||Limited2||X|
|Independent counselor/ mid-level provider||12||Limited5||Limited6||X||X|
|Nursing home/home health||3||Limited8||X|
|Institutional criminal justice||NA||Limited3||X9|
|Measure of demand that can be derived from the indicated source||Number of office visits||Number of outpatient hospital or ER visits||Hospital stays||Point-in-time client count||Number of services and expenditures||Admissions to SUD treatment facilities||Persons accessing any treatment|
X = Indicates the service type is represented in the indicated data source.
|TABLE I.2. Summary of Supply-Side Data Sources|
|Data Source||Ability to Identify SUD Treatment Providers||Key Consideration for Data Collection and Analysis|
|American Community Survey/OES||Categories not specific to SUD treatment providers||Data on employment is reported annually.|
|Certifying organizations (IC&RC and NAADAC)||Providers not certified by IC&RC and NAADAC would be excluded. Certification requirements vary by state. Medical staff and many support workers not included.||May require data collection from certification/licensing boards in 50 states. Data collection varies across boards.|
|Claims data||Providers serving non-Medicare/non-Medicaid populations are excluded from CMS administrative claims. Providers serving managed care enrollees are also often excluded from CMS administrative data. These populations excluded from CMS administrative data represent more than two-thirds of SUD treatment. Data from SDI Health representing privately insured patients is likely to exclude non-prescribers, such as counselors, and services provided under grant funding.||Substantial effort to evaluate completeness and process data. SDI Health data must be purchased.|
|NAMCS||Non-physician providers excluded. Physician weight can be used to identify count of physicians with specific characteristics.||Multiple years of data will need to be pooled to identify a sufficient sample of SUD treatment services.|
|National Plan and Provider Enumeration System||Only professionals with NPIs are included. Professionals are not required to obtain NPIs if they do not prescribe or bill electronically. Preliminary estimates suggest few SUD treatment professionals are identifiable based on the provider taxonomy classifications for addiction medicine.||A list of providers who self-identify in addiction specialties can be identified. However providers are likely to select more general specialty categories such as psychiatrist, internal medicine, or social worker rather than categories specific to addiction.|
|N-SSATS||Workforce questions not included on survey. Supplemental questions were added to 2016 survey for this project. Providers in non-specialty settings, facilities located in jails or prisons, facilities serving only incarcerated clients, and solo practitioners (unless specifically requested by the state) are excluded.||Represents majority of SUD treatment spending nationally. All providers in universe are surveyed. State-level analysis is feasible. Survey is at facility-level.|
|State licensure and certification data||Only state certified or licensed professionals included. Types of professionals that require certification and licensure will vary by state. Medical professionals and professionals working under supervision may be excluded. Sixteen percent of clinical directors are licensed in a field other than substance abuse counseling. Forty-six percent of direct care staff are not licensed.||Significant effort to negotiate data sharing agreement with 50 states. Data will not be in consistent format nor include a consistent set of professionals.|
3. Expert Interviews
To supplement the literature review, we conducted three expert interviews in May and June of 2015. The first interview was with the executive director of the International Certification and Reciprocity Consortium (IC&RC), an organization that develops standards and exams that local boards across the country use for credentialing and licensing. The director responded to our interview questions orally and also provided written responses to the questions on our interview guide from local board staff in Louisiana, Minnesota, North Carolina, and Ohio. The second interview was with the executive director of the National Association for Alcoholism and Drug Abuse Counselors (NAADAC), an association for professionals in the SUD treatment workforce, which also creates exams for certifications. The third interview was with the senior vice president of Public Policy and Practice Improvement with the National Council for Behavioral Health, an association that represents behavioral health provider organizations.
The interviews with the IC&RC and NAADAC representatives addressed addiction provider certification trends, training program trends, state requirements for licensing and certification, and recent changes in the workforce associated with the ACA, MHPAEA and any other observed changes. The interview with the National Council for Behavioral Health representative addressed:
Providers' experiences related to implementation of the ACA and MHPAEA.
How providers have adapted to the availability of expanded Medicaid and private insurance coverage for SUD treatment.
Barriers providers have identified to using this insurance coverage to support patient treatment.
The most pressing concerns for providers related to training programs for SUD treatment professionals, ability to hire qualified staff for open positions, and ability to retain current staff.
State-level differences in licensing/credentialing policies or professional certification requirements that have an important impact on the availability of SUD treatment programs or on program staffing patterns.
The experts were not able to address all topics identified, and had no data available to support response to most questions and therefore could provide only anecdotal information. IC&RC and NAADAC representatives indicated that state board representatives might have information to address particular questions, but this information was not passed on to the national organization.
C. Definition of Substance Use Disorder Treatment Demand
We assess SUD treatment demand under this study with two alternative metrics. First, we used a market-based definition defining demand as observed service use. This definition is useful for identifying the status quo and as a baseline for assessing how demand may change in the future. However, policymakers are keenly interested in the level of unmet need for services. Addressing this question requires an understanding of the prevalence of the illness requiring treatment.
Ideally, market demand would be measured by counting individual units of service used by type of care. However, the available data for this study are limited to information on the number of individuals accessing particular service types within a particular care setting (for example, the number of individuals receiving short-term residential treatment or intensive outpatient services) and the total expenditures for SUD treatment services by settings of care.6
We measure need for treatment based on the prevalence of SUDs. The NSDUH survey provides an excellent resource for identifying individuals who meet diagnostic criteria for abuse or dependence on alcohol or illicit drugs. However, the treatment needs of individuals who meet diagnostic criteria for an SUD vary. Some researchers have noted that standard diagnostic criteria for SUDs may inflate estimates of treatment need. They have proposed alternative concepts which narrow the definition of need. For example, Wakefield and Schmitz (2015) have proposed narrowing the diagnostic criteria for SUDs to require dysfunction and harm (Wakefield and Schmitz 2015). Others have suggested treatment need may be defined in relationship to the individual's own perception of need or based on a disability associated with the SUD (Mechanic 2003). Identifying the populations meeting alternative definitions of treatment need is beyond the scope of the current study. Thus, we present estimates of trends in prevalence of SUDs based on NSDUH, as these trends are likely to parallel trends in the number of individuals who might benefit from treatment.
D. Definition of Substance Use Disorder Treatment Supply
The definition of an SUD professional used in this study is based on three dimensions, as described below.
Engagement in SUD Treatment. The goal of the current study is to assess the adequacy of the workforce to meet demand for SUD treatment services. Only professionals who are actively engaged in the workforce and available to meet this demand will be counted as part of the current workforce. There is considerable turnover in the SUD treatment workforce. Many individuals who have the appropriate education, training, or credentials to provide these services are not currently employed in the SUD treatment field--they are engaged in other professional activities and will not be counted in estimates of the current workforce.
Settings and Nature of Treatment Provided. SUD treatment is provided in specialty and non-specialty settings. This study discusses care in specialty and non-specialty settings. For the NSDUH survey (SAMHSA 2014b) specialty treatment is defined as treatment received at hospitals (inpatient only), drug or alcohol rehabilitation facilities (inpatient or outpatient), or mental health centers. Treatment at an emergency room, private doctor's office, self-help group, prison or jail, or hospital as an outpatient are considered non-specialty treatment. In this report we consider all services reported in N-SSATS specialty treatment.
Education and Experience. Health workforce studies often require that individuals have a specific set of credentials to be counted as part of the workforce. However, because there is limited knowledge about current staffing patterns at SUD treatment organizations, this study takes a broad view of the field requiring only that an individual be actively providing or supporting SUD treatment.