Availability of Mental Health Services Prior to Health Care Reform Insurance Expansions

The expansion of Medicaid and implementation of health insurance exchanges (HIEs) under the Affordable Care Act (ACA), which took effect January 1, 2014, was expected to produce substantial increases in demand for mental health and substance use services. An estimated 52 million individuals have gained mental health insurance coverage since then (1), including 35 million individuals who gained mental health coverage through the parity protections extended to individual and small-group insurance markets (2). These newly insured populations are expected to have higher rates of mental health and substance use treatment needs compared with the population in general (3).

For policy and services planning purposes, it will be important to monitor access to mental health and substance use treatment services to quantify potential workforce and treatment supply constraints and needs. Such constraints may result in limited access to services for newly insured patients. Because a broad range of evidence-based treatments has been shown to be effective in managing mental and addictive disorders and to significantly decrease the illness burden (4), ensuring access to these services is consequential. Although the ACA is designed to improve access to treatment by expanding insurance coverage, improving services delivery models, and promoting payment and other quality improvement and health information technology reforms, the extent to which the nation’s health care systems will be able to meet the expected increase in demand for services—particularly for individuals with mental and substance use disorders—is unclear.

Extensive research has examined specific factors affecting mental health treatment access and identified barriers to care, including attitudinal, financial, and structural barriers (5–8). Studies that have sought to quantify problems related to treatment access have used different methodologies, including epidemiologic studies and claims analyses. Studies of treatment availability have primarily used waiting lists and analyses of disproportionate geographic distribution of the mental health workforce and of treatment facilities. Findings from these studies have consistently documented treatment access and availability problems for individuals with mental and substance use disorders (9–16).

This study was designed to complement prior research by examining systematically collected data from a large probability sample of U.S. psychiatrists to assess their perceptions of mental health treatment availability and accessibility in 2013, the year before implementation of the ACA Medicaid expansions and HIEs. Understanding and documenting treatment gaps are important for policy and services delivery planning to highlight constraints in the availability of specific services and to help ensure that individuals with mental illnesses receive needed treatments.

The primary aims of this study were to assess psychiatrists’ perceptions of the availability of specific treatments and services for mental and substance use disorders and of their ability to spend enough time with patients and provide a sufficient number of visits to meet patients’ clinical needs. The study also identified sociodemographic and practice attributes of psychiatrists who were more likely to report treatment access problems.

Methods

This study used data collected from the Study of Psychiatric Practice Under Health Care Reform, a large probability sample of psychiatrists in the United States. This study was fielded from September through December 2013, before the January 1, 2014, implementation of the ACA insurance expansions. A total of 2,800 physicians who self-identified as psychiatrists and who had not been randomly selected for participation in a survey conducted by the American Psychiatric Association in the past 24 months were randomly selected from the September 2013 release of the Physician Masterfile of the American Medical Association (AMA). Physicians were excluded if they resided outside the United States, if they were psychiatry residents, and if direct patient care was not listed as their type of practice. Persons with undeliverable addresses, who were deceased, or who were incorrectly identified in the AMA Masterfile as psychiatrists (N=185) were excluded. Responses were obtained from 1,188 of the 2,615 psychiatrists in the remaining sample (45%). Of those, 89 were no longer practicing psychiatry or were no longer involved in direct patient care, resulting in a final study population of 1,099 self-identified, currently practicing psychiatrists who were involved in direct patient care.

A four-page data collection instrument was mailed to the target sample with a $50 gift card to increase response. Three survey mailings were implemented at one-month intervals, each followed by a reminder postcard to increase response. All study procedures were approved by the American Psychiatric Foundation Institutional Review Board.

Using a list of ten mental health and substance use treatments and social services (no definitions were provided), participating psychiatrists were asked whether they had any problems providing or finding a referral source of care for each type of service for any individuals during the past 30 days and to indicate the number of individuals who were affected. They were also asked to indicate the number of patients whom they had treated during the last typical workweek for whom they did not have enough time during the patient visit and for whom they were not able to provide the number of visits to meet the patient’s clinical needs.

Weighted analyses were conducted using SUDAAN statistical software to adjust for the sample design, nonresponse, and physician caseload size to provide estimates that could be generalized to the target population (17). Chi square tests (for categorical variables) and Student’s t tests (for continuous variables) examined the associations between physicians’ characteristics and the types and quantity of treatment availability problems. Because of multiple comparisons, a p value of .01 was used to indicate statistical significance.

Results

Psychiatrist Sample

The psychiatrists who participated in the study were primarily male (65%), non-Hispanic whites (82%) (Table 1). A majority were general psychiatrists (75%); 21% reported specializing in child and adolescent psychiatry. Psychiatrists’ patients were most commonly treated in solo office settings (34%) and outpatient public clinics (18%). The primary sources of payment used by patients were private commercial insurance (30%) and self-pay (22%); 43% used public sources. During their last typical workweek, psychiatrists reported treating a mean of 45 patients and spending a mean of 34 hours in direct patient care.

Compared with the sampling frame (N=30,887), respondents were more likely to be non-Hispanic whites (82% versus 67% [N=20,598], p<.001) and were more likely to report specializing in child and adolescent psychiatry (21% versus 17% [N=5,329], p<.05).

Mental Health and Substance Use Treatment Availability

Thirty percent or more of the psychiatrists reported being unable to provide or find a source for clinically or socially indicated psychotherapy (36%), housing (32%), supported employment (31%), case management or assertive community treatment (ACT) (30%), and substance use treatment (30%) in the past 30 days (Table 2). Approximately one in five psychiatrists reported being unable to provide or find a source of care for inpatient treatment (21%), psychosocial rehabilitation (21%), general medical care (20%), pharmacologic treatment (19%), and child and adolescent assessment and treatment (17%) in the past 30 days. In addition, 14% of psychiatrists reported being unable to provide or find sources of care for geriatric assessment and treatment.

Psychiatrists reported being unable to provide or find a referral source for a mean of 2.7 (95% confidence interval=2.5–2.9) services in the past 30 days. A majority (N=813, 74%) reported being unable to provide or find a source for at least one service, 62% (N=681) for two or more services, 47% (N=517) for three or more, 34% (N=374) for four or more, and 23% (N=253) for five or more. The mean number of patients for whom psychiatrists were unable to provide or arrange for each of these services in the past 30 days ranged from 5.0 to 13.6, depending on the service (Table 2).

Ability to Spend Sufficient Time and Provide Enough Visits to Meet Patients’ Needs

Approximately half of all psychiatrists (52%) reported feeling that they did not have sufficient time to treat a patient during their last typical workweek (Table 3). Psychiatrists reported this concern for a mean of 12.8 patients, representing 28% of their patient caseloads. About 40% of all psychiatrists reported being unable to provide a sufficient number of visits to meet a patient’s clinical needs in their last typical workweek. Psychiatrists reported this concern for a mean of 10.6 patients, affecting 24% of their caseloads.

Psychiatrist Factors Associated With Treatment Availability Problems

Psychiatrists who reported a greater number of treatment availability problems were more likely to be younger than age 40 (versus ≥60) and less likely to be non-Hispanic white (versus other race-ethnicity) (Table 4). They were also more likely to see a majority of their patients in public outpatient clinics (versus solo practices); treat more than 44 patients per week (versus ≤44); provide more than 40 hours of direct patient care per week (versus ≤40); and primarily treat Medicaid patients (versus patients with private insurance). Time and visit constraints were commonly reported across solo, group, and public outpatient settings (Table 3). Psychiatrists who saw a majority of their patients in public outpatient clinics, treated more than 44 patients, or provided more than 40 hours of direct patient care in a typical week reported not having sufficient time nor providing sufficient visits to meet clinical needs for a greater number of patients during the last typical workweek (Table 4).

Discussion

Mental Health and Substance Use Services Availability

Substantial proportions of psychiatrists reported being unable to provide or find a referral source of care for individuals needing each of the ten services examined. These findings are consistent with prior research indicating a constrained supply of mental health and substance use services (9–11,18,19), affecting sizable numbers of patients seeking treatment from psychiatrists. Psychiatrists’ perceptions of gaps in access to general medical care were also consistent with prior research (20). These perceptions are concerning in light of the significant excess morbidity and mortality among individuals with mental illnesses compared with the general population.

This study is notable in highlighting reports of constrained availability of even the most common mental health treatments, such as psychotherapy and medications. Although psychotherapy and medications remain first-line or adjunctive treatments for most mental illnesses (4), mounting evidence supports the superior effectiveness of a combination of modalities for a number of conditions (21). Many individuals who could benefit from combined treatments may be receiving only one type of treatment because of availability or access constraints.

Treatment availability problems may have increased since our data were collected, given that the data were collected before the ACA insurance expansions, which have increased the number of newly insured individuals seeking services. Consequently, the considerable social, economic, and human costs and burdens associated with inadequate mental health treatment access necessitate several careful considerations of health and social policy, financing, and services delivery approaches that could ameliorate the identified treatment gaps.

First, although more efficient delivery, coordination, and team-based management of care may facilitate improved access to care (22), the scope and magnitude of reported treatment availability problems in our study and in prior health care workforce studies highlight the need for increased investments in mental health, substance use, and primary care services and workforce development and training. Physician, nursing, and behavioral health workforce shortages and maldistribution problems have been well documented and are expected to worsen as a result of the aging of the U.S. population and health care workforce (19,23). Although the ACA included provisions to address the size, composition, and geographic distribution of the primary care and behavioral health workforce, these provisions included limited measures to address the shortage of psychiatrists. In addition, the extent to which these provisions will address workforce supply is unknown because they have not been fully implemented, they are temporary, and they rely on discretionary funding (24).

Second, ensuring that the full range of evidence-based treatments is covered under state Medicaid programs and HIE plan offerings will be essential for availability of and access to mental health services. It remains doubtful whether coverage for essential health benefits for persons gaining health insurance will include services not traditionally covered by commercial “benchmark plans” and only sometimes covered by Medicaid programs, such as supported employment, supported housing, psychosocial rehabilitation, and case management or ACT (3,9). The high rates of reported availability problems for evidence-based services is concerning, given that these services provide the foundation for effective, coordinated early intervention specialty care models such as the Recovery After an Initial Schizophrenia Episode model (25) for individuals with serious mental illness (26). Continued advocacy for funding and policies that facilitate access to these services is needed.

The actual levels of services covered by plans in the HIEs and by Medicaid programs, in terms of numbers of visits, inpatient days, and range of psychosocial and psychopharmacologic treatments offered, also remain uncertain. There are concerns that the services offered may not meet patients’ clinical needs, particularly for patients with severe mental illnesses, who are disproportionately represented in ACA expansion populations. In addition to concerns about quantitative coverage limits, concerns remain regarding the effect of nonquantitative treatment limits on treatment access, despite federal mental health parity protections under the ACA. Such limits include the medical management of benefits through medical necessity criteria, prior authorizations, and formulary designs, in addition to provider network and reimbursement adequacy issues (27,28).

Third, sound financing and reimbursement policies are essential to promoting appropriate delivery and organization of the levels and types of needed mental health and substance use services. Health homes, accountable care organizations, and the myriad health care delivery and payment demonstrations funded by the Centers for Medicare and Medicaid Services’ Center for Innovation hold promise for improving access to the fuller range of services needed by individuals with mental illnesses. However, the overall levels and structure of funding with health plans and health care providers, and the quality performance monitoring and financial incentive parameters implemented, will have major implications for the extent to which improved access to mental health services is realized. Sufficient levels of funding and financial incentives and protections targeted to individuals with mental health and substance use disorders will be essential to ensure availability of and access to care. This includes clinically rigorous “merit-based” or “value-based” payment methods that reflect the full range of evidence-based treatments and provider reimbursement levels and administrative terms sufficient to attract high-quality clinicians. This may prove challenging in an era of constrained resources and increasing numbers of insured individuals. Further, onerous quality-reporting demands and the risk of financial penalties might deter clinicians’ participation in these arrangements.

Psychiatrists’ Ability to Provide Sufficient Time and Visits

For a number of reasons, it would be helpful to better understand the validity and nature of psychiatrists’ perceived time and visit constraints and the potential effects of these constraints on quality and outcomes of care. These constraints may lead to insufficient patient assessments or insufficient provision of psychosocial treatments and education, resulting in poorer care. They could adversely affect best practices, such as measurement-based care and ongoing monitoring and adjustment of medications and dosing to optimize outcomes and minimize side effects (4).

Given both the medical and psychosocial complexity of many patients’ experiences with mental illnesses, it may be particularly challenging for psychiatrists to address the needs of severely ill patients during brief encounters or a limited number of encounters. New ACA models of integrated, collaborative care may help address these challenges by encouraging psychiatrists to step into leadership roles and work more closely with general medical and behavioral health providers to provide patient-centered, team-based care (29). Psychiatrists may play an increasing role in providing basic prevention and treatment for common general medical conditions, particularly for vulnerable patients with serious mental illness (30). Psychiatric patients, in particular, face constrained access to primary care services because of the shortage of primary care practitioners (31), and they frequently do not have a designated primary care provider (20). Because psychiatrists are trained in general medical and behavioral health care, they regularly evaluate and manage patients with general medical and psychiatric issues. This often requires that they spend more time with a patient compared with primary care providers, consistent with psychiatrists’ reports of patient time constraints.

Given the significantly higher treatment availability problems reported by psychiatrists who primarily treat Medicaid patients and practice in group or public clinics, exploring the nature and effect of constraints in these populations warrants more attention. Prospectively identifying attributes of patients associated with time and visit constraints through patient screening could facilitate scheduling additional time or visits as needed. Understanding how health plan coverage and management issues, including limits on the length or number of visits, are related to these constraints may be informative if there are associated effects on treatment quality and outcomes.

Strengths and Limitations of the Study

The primary strength of this study was the use of a large probability sample of U.S. psychiatrists, who represent the primary physician discipline specializing in treating mental and addictive disorders, treating an estimated 20% of individuals who receive mental health care in the United States (32). Because of their training as physicians, psychiatrists also are more likely to treat severely ill patients with complex, comorbid conditions requiring a range of pharmacologic and psychosocial treatments.

Our data are subject to psychiatrist response, recall, and social desirability biases. The exclusive reliance on physicians’ reports without external or psychometric validation of measures represents a significant limitation. Psychiatrists, particularly those who practice in solo and group office settings—who comprised half the sample—may be unaware of services available in their communities. As a result, they may overestimate treatment availability problems. Furthermore, the study’s global treatment availability measures did not distinguish between treatment availability and accessibility problems. We did not assess whether respondents tried to find services or any specific access problems encountered while doing so. In addition, even when services are available in the community, they may not be accessible to specific individuals or groups, such as low-income, uninsured, or publicly insured patients; children; non–English-speaking individuals; and active users of illicit substances.

Simple random sampling resulted in a final study sample primarily reflecting psychiatrists from more populous states, such as California, New York, Massachusetts, Florida, and Pennsylvania. Although psychiatrists from 48 states participated in the study, the sample included no psychiatrists from Delaware and West Virginia and only a few from Virginia, Wyoming, North and South Dakota, Alaska, Nevada, and Montana.

Conclusions

Findings from this study indicate that psychiatrists report significant constraints in the availability of a range of mental health, substance use, and general medical services essential for individuals with mental illnesses. Given the magnitude of these reported problems, significant enhancements of the mental health and substance use treatment workforce and services delivery infrastructure may be required in order for the ACA to realize its promise of improved health, better health care delivery, and lower costs.