Assessing Patient Characteristics and Treatment Patterns Among Psychiatrists Who Do Not Accept Private Insurance

Abstract

Objective:

Privately insured individuals frequently use of out-of-network psychiatrists. Yet, whether treatment provided by psychiatrists who do not accept private insurance differs from treatment provided by those who do has not been studied. The investigators describe provider characteristics, patient characteristics, and treatment patterns among psychiatrists who do not accept new patients with private insurance.

Methods:

Data for this study come from the National Ambulatory Medical Care Survey (2011–2014), a nationally representative annual cross-sectional survey of physicians providing ambulatory care. Responses of psychiatrists who report accepting any new patients were examined(N=440), representing 7634 visits.

Results:

Compared to psychiatrists accepting privately insured patients, psychiatrists not accepting privately insured patients had fewer visits with patients with serious mental illness (42% versus 53%; p=.016). These psychiatrists had a higher proportion of visits of duration greater than 30 minutes (48% versus 34%; p=.026), and their patients were more likely to have had 10 or more visits in the past 12 months (41% versus 28%; p=.013). There were no differences in the proportion of visits in which treatment included psychotherapy (48% vs 44%; p=.518).

Conclusions:

Although psychiatrists not accepting patients with private insurance are less likely to treat patients with serious mental illness, patients of these psychiatrists were more likely to have longer visits and a relatively high number of visits in the past year. The low rate of acceptance of insurance among psychiatrists may have the greatest effect among those most in need of services.

Among the 43.4 million people with a mental illness in the US, only about 43% received treatment in the past year (1). One often-cited barrier to treatment is access to a mental health specialty provider, in part because psychiatrists have been less likely than other physicians to participate in private (commercial) insurer networks (2). The most recent research examining psychiatrists’ participation in plan networks examined data through 2010 and found that participation was declining significantly, with only about 55% of psychiatrists accepting new patients with private insurance by 2009–2010 (3). This compared to rates of about 90% for other physician specialties. Although privately insured patients obtain mental health care from out-of-network providers at high rates (4), little is known about the types of patients treated by psychiatrists who do not accept insurance and whether the treatment they deliver differs from that delivered by in-network psychiatrists.

Related to patient characteristics, differences in demand for specialty mental health treatment may be relevant. Privately insured patients willing to pay the higher costs associated with out-of-network care may have less serious diagnoses, in part because they are more likely to have stable employment and higher income. Conversely, those with the most serious symptoms may be eager to see a provider quickly, even if it comes at the significantly higher cost associated with an out-of-network psychiatrist. Psychiatrists who prefer to treat individuals with less serious diagnoses, perhaps to reduce off-hours calls and appointments (particularly for psychiatrists in solo practice), may decline to accept private insurance to allow for more discretion over the types of patients joining their patient panel.

Psychiatrists who do not accept private insurance may also provide different treatments. If insurers set relatively low payment rates for psychotherapy to encourage patients be treated with lower cost treatments (i.e., medication), and patients are willing to pay out-of-network rates for psychotherapy, psychiatrists that have a preference for providing psychotherapy or believe psychotherapy (perhaps in combination with medication treatment) is the most effective treatment may choose not to accept private insurance. Moreover, not accepting private insurance may allow some psychiatrists to avoid strict and time-consuming medical necessity reviews that are triggered by longer treatment episodes.

Generally, mental health providers cite low reimbursement, excessive paperwork, and late or incorrect payments as reasons for lack of private insurance network participation (5,6,7). Implementation of the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity Act of 2008 (MHPAEA) may have addressed some of these barriers to psychiatrists’ participation in networks. Under MHPAEA many private insurance plans are required to use a similar process to determine the management rules for behavioral health and medical/surgical care, potentially reducing administrative requirements on mental health providers (8). Similarly, MHPAEA prohibits most private insurance plans from using different standards for choosing or reimbursing in-network mental health providers than those used for choosing or reimbursing in-network medical/surgical providers. There is some evidence that plans responded to these requirements by expanding their mental health provider networks. One large 2011 nationally representative survey of plans reported that four out of five plans increased the number of providers in their behavioral health network post MHPAEA (9).

In the present study, we describe provider characteristics, patient characteristics, and treatment patterns among psychiatrists who do not accept new patients with private insurance (2011–2014).

Methods

Data Source

We examine the 2011–2014 National Ambulatory Medical Care Survey (NAMCS) conducted by the Centers for Disease Control and Prevention’s National Center for Health Statistics, an annual survey of office-based physicians. The sampling frame consists of physicians in the masterfiles of the American Medical Association and the American Osteopathic Association who are primarily in office-based practice and principally engaged in patient care activities. Visits to community health centers are not included. In all cases, we exclude from our sample any providers reporting not accepting any new patients. Over the time period studied, 90% of psychiatrists and 96% of all other specialties were accepting new patients. Each provider reports information about several randomly chosen patient encounters.

Measures

Our main outcome measure is whether the provider self-reports accepting patients with private insurance. Each provider in the survey is asked, “If you currently accept “new” patients into your practice(s), which of the following types of payment do you accept?” If a provider is not accepting new patients, information about payment types accepted is not requested. Thus, in this study we focus only on providers that report they are accepting new patients. Following methods originally used by Bishop et al., we create a measure of acceptance of new patients with non-capitated private insurance (3). Other psychiatrist practice characteristics we consider include region, whether the psychiatrist practices in an MSA, is in solo practice, or has evening or weekend availability. Starting in 2011–2013, providers were also asked about their electronic health record capabilities. We consider whether the practice submits claims electronically, meets HHS meaningful use criteria, uses clinical notes recording, or uses secure messaging exchanges with patients.

Patient encounter or visit data considered includes whether the patient was aged 19 or under, female, a new patient, had two or more chronic conditions, had a diagnosis indicating serious mental illness (serious mental illness; ICD-9-CM diagnoses 295.xx-298.xx), or a diagnosis indicating substance use disorder (ICD-9-CM diagnoses 291.xx-292.xx and 303.xx-305.xx, except 305.1 Tobacco Use Disorder). Related to treatments received, we considered whether the visit included psychotherapy, whether medication was ordered or provided during the visit, whether the duration of visit was 30 minutes or greater, and whether the patient had 10 or more visits in the past 12 months. NAMCS defines psychotherapy as “all treatment involving the intentional use of verbal techniques to explore or alter the patient’s emotional life in order to effect symptom reduction or behavior change” and distinguishes psychotherapy from “all other mental health counseling” (10). Duration of visit includes only the amount of time the physician spent in face-to-face contact with the patient.

Statistical Analyses

We consider whether psychiatrists accepting new non-capitated private insurance and those not accepting private insurance differ with respect to characteristics of provider practices and characteristics of patients treated by these providers. In all analyses, provider or visit survey weights are used to make results nationally representative. The design variables provided by NCHS are used and we employ an ultimate cluster design for estimating variance. This study was declared exempt from review by the Yale Human Investigation Committee. We use Stata 14.2 for all analyses.

Results

Our final NAMCS sample includes 440 psychiatrists responding in 2011–2014, with data for 7634 visits, representing 119,605,827 patient visits nationally. Among psychiatrists, 65% were accepting new privately insured patients. This number was significantly higher for other physician specialists -- 89% (p<.01).

Psychiatrists’ practice characteristics

We considered differences in psychiatrists’ practice characteristics in Table 1. Psychiatrists not accepting new privately insured patients were more likely to report being in solo practice (75% versus 63%; p=.037). For both types of providers, more than 97% of practices were in an Metropolitan Statistical Area. Psychiatrists not accepting private insurance were also less likely to accept new patients with Medicare (39% versus 75%; p<.001) or Medicaid (31% versus 52%; p=.001). There was a significantly less likelihood that a practice submits claims electronically (46% versus 77%; p<.001), uses clinical note recording (44% versus 62%; p=.046), and that the practice uses secure messaging exchanges with patients (8% versus 19%; p<.05) among those not accepting new patients with private insurance.

Table 1:

Characteristics of Psychiatrist Practices (NAMCS)

	Psychiatrists Accepting New Patients
	All (N=440)	Does not accept private insurance (N=141)	Accepts private insurance (N=299)	p-value for test of group difference
Practice characteristics (2011–2014)
Region (%)				0.113
Northeast	27	35	25
Midwest	17	10	19
South	28	26	32
West	28	29	24
Practice is in Metropolitan Statistical Area (%)	98	99	97	0.042
Solo Practice (%)	64	75	63	0.037
Evening and/or Weekend Availability (%)	39	47	36	0.113
Accepting Medicare Patients (%)	63	39	75	<.001
Accepting Medicaid Patients (%)	46	31	52	0.001
Electronic health record capabilities (2013–2014)
Practice Submits Claims Electronically (%)	68	46	77	0.001
Practice Meets HHS Meaningful Use Criteria (%)	90	88	90	0.761
Practice Uses Clinical Notes Recording (%)	57	44	62	0.046
Practice Uses Secure Messaging Exchanges with Patients (%)	17	8	19	0.050

^*Survey-weighted percentages are based on the sample that was surveyed, which was limited to psychiatrists who reported accepting new patients. P-value compares psychiatrists not accepting new patients with non-capitated private insurance to psychiatrists accepting new patients with non-capitated private insurance. Physician characteristics are weighted with the NAMCS physician weight. Electronic health record capabilities include data from 2013–2014 because some measures were not included in 2011 and 2012 surveys.

Visit characteristics

Psychiatrists not accepting new privately insured patients were significantly less likely to see individuals with serious mental illness (42% versus 53%; p=.016), although there were no significant differences in the proportion of visits by patients with at least two general medical conditions (18% versus 22%; p=.342) or by whether the patient had a substance use disorder diagnosis (11% versus 9%; p=.371). Psychiatrists not accepting new privately insured patients were marginally significantly less likely to treat individuals ages 19 and under (11% versus 17%; p=.075; see table 2).

Table 2:

Characteristics of Psychiatrists’ Patient Visits (2011–2014)

Visit Characteristics (2011–2014)	All Visits with Psychiatrists Accepting New Patients
	All visits (N=7634)	Does not accept private insurance (N=2243)	Accepts private insurance (N=5391)	p-value for test of group difference
Patient Characteristics (%) Age 19 and under		11	17	0.075
	16
Female	56	56	57	0.831
New patient	9	7	9	0.235
Patient has 2 or more chronic conditions	20	18	22	0.342
Patient has serious mental illness	50	42	53	0.016
Patient has Substance Use Disorder (SUD)	9.5	11	9	0.371
Treatment Characteristics
Visit included psychotherapy	43	48	44	0.518
Any prescription or non-prescription drugs ordered or provided during this visit	87	84	88	0.244
Visit lasted 30 minutes or more	39	48	34	0.026
Patient had 10 or more visits in the past 12 months	32	41	28	0.013

Survey-weighted percentages of patient encounters are based on the sample that was surveyed, which was limited to visits with psychiatrists who reported accepting new patients. Patient characteristics are weighted with the NAMCS patient weight.

Related to treatment provided during the visit, there was no significant difference in the proportion of visits that included treatment with psychotherapy (48% versus 44%; p=.518) or whether any prescription or non-prescription drugs were provided during the visit (84% versus 88%; p=.244). Visits among psychiatrists not accepting new privately insured patients were significantly more likely to last 30 minutes or longer (48% versus 34%; p=.026), and to be with patients with 10 or more visits in the past 12 months (41% versus 28%; p=.013).

Discussion

This study is the first to look at patient characteristics and treatment patterns associated with whether a psychiatrist is accepting new privately insured patients. We find that psychiatrists who accept private insurance may be seeing patients with more complex treatment needs - psychiatrists who accept insurance have approximately 25 percent more visits involving a patient with a diagnosis indicating serious mental illness. This difference in patient panel may be due to patient demand or provider behavior. If patients with more serious diagnoses anticipate needing more services, the cost of seeing out-of-network psychiatrists may be prohibitively expensive. In addition, these patients may be more likely to benefit from ancillary services such as case management, which may be more available among psychiatrists who accept private insurance. Provider behavior may also play a role if providers that choose not to participate in private insurance networks are more willing to accept new patients with less serious diagnoses (i.e. “cream skimming” the easy-to-treat patients).

Despite seeing more patients with serious mental illness, psychiatrists who accept insurance on average spent less time with their patients and saw them fewer times than psychiatrists who do not accept insurance. We found that visits were more likely to be 30 minutes or more among psychiatrists not accepting private insurance, and that their patients were more likely to have had 10 or more visits in the past 12 months, suggesting more intense treatment episodes. Insurers may be managing care, and be more likely to require a justification for longer visits or treatment episodes of exceptionally long duration, resulting in the less intense care provided among in-network providers. Psychiatrists not participating in private insurance networks may feel less pressure to end treatment with a patient, particularly if they receive fewer requests for appointments from new patients, given their higher out-of-pocket cost.

Anecdotal reports suggest that some psychiatrists do not accept new patients with private insurance due to a preference for providing psychotherapy (compared with medication management), which at longer durations may be less likely to be approved by plans. That we found no differences in the proportion of visits that included psychotherapy suggests other important reasons for not accepting patients with private insurance.

Practice characteristics also differed. We found that psychiatrists who accept new patients with private insurance are more likely to have electronic medical record systems with greater capability, which may reduce billing costs or errors. Psychiatrists who do not accept insurance are more likely to be in solo practice, and may not be able to afford the overhead or complex implementation of an electronic medical record. These systems may lead to higher quality care if access to information about general medical conditions is more readily available. Overall, psychiatrists had fewer electronic medical record capabilities than other specialists. In our sample, 62% of psychiatrists could submit claims electronically in 2014; one NAMCS report indicated when considering all specialties, 89% could submit claims electronically (11).

Policy observers have suggested that two important reasons psychiatrists may not accept new privately insured patients are low overall reimbursement rates and administrative hassles associated with reimbursement by private insurers. Information on whether these psychiatrists accept new patients with Medicare coverage may be informative on the relative importance of these two concerns. Medicare historically has fewer administrative hurdles to payment once care is delivered. Even though administrative hurdles under private insurance may have been reduced somewhat due to MHPAEA, they are still likely higher than under Medicare. We find that about 39% of psychiatrists not accepting new privately insured patients do accept new Medicare patients, suggesting administrative burden may be an important reason for not accepting private insurance for these psychiatrists.

Rates of psychiatrists accepting new privately insured patients remain quite low compared to other specialties, justifying ongoing concerns about access to mental health care for the privately insured. Lack of acceptance by psychiatrists may impede or delay receipt of care, particularly if provider directories are not accurate (12). It may also lead to additional financial burden on patients due to higher cost sharing for out-of-network care as well as balance billing (4). Although not generalizable, a patient and family survey conducted by the National Alliance on Mental Illness found that 30–32 percent of respondents reported having difficulty finding a mental health specialist who would take their insurance (13). Unlike many prior state parity laws, the MHPAEA required plans that offer an out-of-network benefit for mental health have a benefit equivalent to out-of-network medical/surgical care (including copayment, coinsurance, out-of-pocket maximums and deductibles). Critics have noted the difficulty enforcing requirements in MHPAEA and there have been numerous complaints by patients about policies not in compliance with MHPAEA (14,15). There has also been significant criticism of lack of transparency of some of the provisions. Perhaps with greater enforcement and disclosure requirements, plan policies may change in such a way as to encourage more psychiatrists to accept private insurance.

Delivery system reforms have led to increased emphasis of the treatment of mental health disorders in primary care, perhaps by a team of providers that include a mental health specialty provider (16). This change may alleviate provider shortages and increase access to mental health care, even in the face of limited participation by psychiatrists in private insurance networks. In addition, other mental health specialty providers, including psychologists and social workers, may provide needed mental health services, particularly psychotherapy. Yet, for individuals with more serious mental illness, access to a psychiatrist may still be critical. Thus, when there is a shortage of psychiatrists, it may be optimal at a population level for some patients, particularly those with less severe anxiety disorders or depression symptoms, to be treated in a primary care setting.

Limitations of the data are important to consider when determining implications of these findings. We only consider psychiatrists because data on other mental health professionals was not available. These data do not include all outpatient visits; hospital based outpatient care or care delivered in community health centers are not included. Although nationally representative, the sample size of psychiatrists practice in each year is relatively small, and some meaningful differences may not be detected. There are well known limitations to survey data such as difficulties accurately determining diagnoses and other patient characteristics or treatments patterns. This may be particularly true in the case of psychotherapy, where treatment that constituted psychotherapy may have been interpreted differently by different psychiatrists. We use recorded diagnosis to identify serious mental illness, which in the absence of information on functional limitations may not accurately categorize some patients. Finally, this study is descriptive in nature, and we are unable to determine what policies such as higher reimbursement or changes to health plan pre-certification or treatment review may ameliorate the problem of low psychiatrist acceptance of new privately insured patients. Likewise, we cannot determine the consequences of this problem for patient outcomes.

Conclusions

Although psychiatrists not accepting patients with private insurance are less likely to treat patients with serious mental illness, patients of these psychiatrists were more likely to have longer visits and a relatively high number of visits in the past year. The low rate of acceptance of insurance among psychiatrists may have the greatest effect among those most in need of services.