Differential Reimbursement of Psychiatric Services by Psychiatrists and Other Medical Providers

In 2013, there were 11.0 million adults ages 18 or older in the United States who perceived an unmet need for mental health care (1). The two most commonly reported reasons for not receiving the needed care were an inability to afford treatment (48%) and not knowing where to go for services (25%) (1).

The shortage of mental health providers contributes to treatment access barriers. More than 75% of counties have what is considered a severe shortage of psychiatrists (2). From 2008 to 2013, the number of active psychiatrists in the United States decreased from 38,857 (one per 7,825 individuals) to 37,296 (one per 8,476 individuals) (3). Furthermore, 59% of psychiatrists are age 55 or older, and more than half of psychiatrists may be retiring in the next decade (3). Compounding the patient access issue is the growing trend for psychiatrists not to participate in insurance networks (4). From 2005 to 2015, the percentage of psychiatrists who accepted private insurance declined from 72% to 55% and was significantly below that of other specialists (89%) (5). Psychiatrists may choose not to participate in insurance networks because of a perception that insurance reimbursement is low (6). Also in specialties that have provider shortages, like psychiatry, the traditional network participation offer of discounted reimbursement in exchange for increased patient volume from the network may be less appealing compared with specialties that do not have shortages. The reason is that physicians in specialties with shortages are likely already experiencing high demand for their services.

Recent changes to the psychiatric Current Procedural Terminology (CPT) codes provide an opportunity to investigate this important issue by testing, for the first time, whether insurers reimburse psychiatrists less than other providers for offering the same services (7). Determining how psychiatrists are reimbursed within the network compared with how other network providers are reimbursed for the same services and how psychiatrists are reimbursed when services are paid out of network can potentially inform our understanding of psychiatrists’ low participation in insurance networks. In addition, examining the out-of-pocket payment implications for patients who receive care out of network may have implications for patient access to care by psychiatrists.

Methods

The study used 2014 data from the Truven Health Analytics MarketScan Commercial Claims and Encounters (MarketScan CCAE) database. The MarketScan CCAE database is created by aggregating data from commercial health plans. It captures the entire claims experience of approximately 30 million individuals with private insurance annually from all 50 states. The data are deidentified, and they capture all insured services received, including any services covered by a behavioral health managed care specialty plan (a “carve-out”). The database also includes a measure of the total amount of the insurance claim submitted by the providers, the amount that was reimbursed by the health plan, and the amount paid out of pocket by the consumer. In addition, the data set has a measure of whether the provider was in a beneficiary’s insurance network or outside that network.

The sample for the analysis was restricted to paid insurance claims for professional office visits for treatment of a patient with a primary mental disorder diagnosis (ICD-9 codes 290–316, which include alcohol and drug dependence). These claims were submitted by employees or their dependents, psychiatrists (including child psychiatry specialists), nonpsychiatrist medical doctors, psychologists, social workers, and psychiatric nurse practitioners. For the category of nonpsychiatrist medical doctors, we included all family practice, internal medicine, and pediatric physicians. The final sample included 3.8 million patients.

To compare reimbursement for the same services by provider type, we identified the outpatient services provided most commonly by each of the five provider types mentioned above during visits by patients with a primary mental disorder diagnosis. The median reimbursement (the combination of the insurance reimbursement and the out-of-pocket payment) received by psychiatrists and nonpsychiatrist medical doctors for delivery of these common behavioral health services was compared. The median was compared because it is influenced by outliers to a lesser degree than the mean. Because reimbursement rates may vary across regions, reimbursement in several large, geographically dispersed metropolitan statistical areas (MSAs) (New York, Los Angeles, Atlanta, and Lincoln, Nebraska) and in all non-MSA areas was also calculated. [A table summarizing reimbursement by locality is available as an online supplement to this article.] The study also evaluated whether psychiatrists receive greater reimbursement for out-of-network services compared with other providers. Finally, we determined whether median patient out-of-pocket payments for the same services differed by provider.

Results

Table 1 identifies the most commonly delivered services for patients with primary mental disorder diagnoses by type of provider. The services provided most commonly by psychiatrists, nonpsychiatrist medical doctors, and psychiatric nurse practitioners during visits by patients with a primary mental disorder diagnosis were evaluation and management (E/M) services corresponding to office visits for presenting problems of low to moderate severity (CPT code 99213) and for presenting problems of moderate to high (99214) severity. They comprised 44.8% of services delivered by psychiatrists, 58.9% of services delivered by nonpsychiatrist medical doctors, and 49.3% of services delivered by psychiatric nurse practitioners. Psychotherapy was also among the most common services provided by psychiatrists and psychiatric nurse practitioners during visits by patients with a primary mental disorder diagnosis, as reflected in codes 90833, 90836, and 90834, but it was not among the most common services delivered by nonpsychiatrist medical doctors. For social workers and psychologists, the most common psychotherapy codes used during visits by patients with a primary mental disorder diagnosis were 90834 and 90837, corresponding to 45-minute and 60-minute psychotherapy visits. These services account for more than 75% of services provided by those two provider types.

Table 2 shows the median total insurance reimbursement, in network and out of network, for the two most common services provided by psychiatrists, other physicians, and psychiatric nurse practitioners during visits by patients with a primary mental disorder diagnosis (E/M codes 99213 and 99214). The median reimbursements for in-network services were lower for psychiatrists than for nonpsychiatrist medical doctors ($66 versus $76 for 99213 and $91 versus $114 for 99214). Translated into percentages, psychiatrists received 13% (99213) and 20% (99214) less in total payments for in-network services compared with other physicians. For out-of-network services, the pattern was reversed, with psychiatrists being reimbursed $100 for CPT code 99213 versus $78 for other physicians and $122 for CPT code 99214 versus $115 for other physicians. This translates into psychiatrists being reimbursed 28% more than other physicians for delivery of services coded 99213 to out-of-network patients and 6% more for services coded 99214. Another way to view the results is that the differential reimbursement between in-network and out-of-network services for the same CPT codes was much greater for psychiatrists than for other physicians. For CPT code 99213, psychiatrists received 52% more in total payment for out-of-network services compared with in-network services, versus a differential of 3% for other physicians. For CPT code 99214, they received 34% more in total payment for out-of-network services compared with in-network services, versus a differential of 1% for other physicians.

Analysis of reimbursement within large MSAs and in all non-MSA areas revealed a similar pattern. The results by region were consistent with the overall differences, with psychiatrists being reimbursed less than other nonpsychiatrist physicians for in-network services with the same procedure codes for patients with a primary mental disorder diagnosis. The only exception to this was Los Angeles, where psychiatrists were reimbursed more than nonpsychiatrist physicians for the same services [see online supplement].

Table 3 shows the differential in the median patient out-of-pocket payment for visits to various providers. For CPT code 99213, the median in-network out-of-pocket payment was the same for psychiatrists, other physicians, and psychiatric nurse practitioners ($20). For code 99214, the median in-network out-of-pocket payment was higher for nonpsychiatrist physicians than psychiatrists ($24 versus $20). Out-of-pocket costs to patients for out-of-network psychiatrist visits were about twice as high as those for in-network visits. For out-of-network services, median out-of-pocket payments for visits to psychiatrists were higher than for visits to nonpsychiatrist physicians, for both code 99213 and code 99214. In addition, Table 3 shows the percentage of visits that were out of network for each of the three types of service providers. The percentage of out-of-network visits for the two procedures was more than twice as high for psychiatrists than for each of the other two provider types.

Table 4 compares the median payments (in network and out of network) among mental health professionals for the most common psychotherapy billing codes of similar duration. Psychiatrists were reimbursed less for psychotherapy add-on services in network than psychologists were for psychotherapy standalone services of similar duration; however, psychiatrists received a higher payment when they provided the psychotherapy as a standalone service (which occurred in only 6.1% of visits). For all three mental health professionals, the median reimbursement for psychotherapy lasting 38 to 52 minutes (90834 for psychologists and social workers and 90834 or 90836 for psychiatrists) was significantly higher for out-of-network claims than for in-network claims (43%, 81%, 70%, and 87% higher for psychiatrists providing stand-alone psychotherapy, for psychiatrists providing a psychotherapy add-on, for psychologists, and for social workers, respectively). The results by region were consistent with those reported in Table 4 [see online supplement].

Table 5 shows the percentage of psychotherapy services that were provided outside the network, according to the bill. For example, 34% of visits to psychiatrists for standalone psychotherapy lasting 38 to 52 minutes (CPT 90834) were provided out of network. The use of out-of-network providers results in patients facing significantly higher out-of-pocket payments for psychotherapy, as shown by the median out-of-pocket payments for in-network and out-of-network providers of psychotherapy services (Table 5). For example, the differential out-of-pocket costs to patients for visiting a psychiatrist out of network versus in network for 38–52 minutes of psychotherapy was $25.

Discussion

In 2014, claims for more than 40% of visits to psychiatrists and nonpsychiatrist medical doctors by commercially insured patients in our sample with a primary mental disorder diagnosis used the same two CPT codes. The in-network reimbursements for these two codes were lower for psychiatrists compared with other types of physicians. Although the analyses do not provide causal estimates about why fewer psychiatrists participate in insurance networks, the findings might help explain why psychiatrists are less likely to participate in insurance networks. The study also demonstrated that when patients receive care from out-of-network providers for services billed with the same common CPT codes, their out-of-pocket payments are higher for psychiatrists than for nonpsychiatrist medical doctors and psychiatric nurse practitioners.

Clear information on health insurance coverage and plan policies can help consumers choose health plans and understand health benefits given their needs and the options available to them (8). Information on the participation of psychiatrists in insurance networks has been demonstrated to be very poor (9–13). Even though insurance companies’ use of physician networks can help contain costs for consumers, it might also result in less than optimal coverage for certain services delivered by particular types of providers, especially if there is a lack of price transparency in the cost of going out of network. Many insurance plans no longer pay usual, customary, and reasonable (UCR) rates, and they are moving more and more toward adopting proprietary or public fee schedules (such as Medicare’s resource-based relative value scale) to set fees for out-of-network services (14). This practice can make it difficult for consumers to know how much they will be reimbursed for treatment from an out-of-network provider, which can affect access (15).

Despite the strength of the data used in this study, the results should be interpreted in light of a few considerations. First, this study was limited to the analysis of a large convenience sample of private insurance claims. The MarketScan CCAE database, however, captures approximately 25% of employer-sponsored health care beneficiaries, and the coverage area is the entire United States. Further insights could be gained by repeating this study by using other private insurance claims databases, as well as Medicaid and Medicare claims. Second, the study did not capture payments for treatment that were not submitted to insurance plans or claims that were submitted and denied. Although there is no published research on the issue, online forums reveal that some consumers may choose not to submit out-of-network claims for partial reimbursement for a variety of reasons, including privacy concerns and a wish to avoid paperwork, or because their insurance does not cover services out of network (16,17). Third, there are additional considerations with regard to physicians’ insurance network participation that could not be measured in this study, such as administrative costs associated with insurance network participation and billing. The shortage of psychiatrists can result in excess demand for their services. The excess demand allows psychiatrists to serve patients who are willing to pay out of pocket, potentially lessening the psychiatrists’ administrative costs. Fourth, some patients are not able to receive services because they cannot find providers who accept their insurance or they cannot afford the out-of-pocket payments (1). This study does not capture services that patients may need but did not receive.

Our study is intended to provide an overview of the issue of payment disparities for mental health services, and, as such, it focuses broadly on commercially insured individuals receiving mental health care. Although it included services provided by child psychiatrists, it does not provide analysis of them specifically. There is a particular shortage of child psychiatrists in the United States (18), and further research is needed to explore the disparities in payment for psychiatrists and other medical doctors serving youths with mental disorders in particular.

Conclusions

The share of psychiatrists who accept insurance is lower than that of nonpsychiatrist physicians and is declining (4). To address this issue, policy makers and payers may need to confirm that psychiatrists and other mental health professionals are equitably incentivized to participate in insurance networks by comparing reimbursement rates for comparable services and diagnoses. Providing consumers with accurate, user-friendly information on the number of psychiatrists and other behavioral health providers participating in insurance networks (and whether those providers are accepting new patients) and the reimbursement amounts for in-network and out-of-network services may also help to improve access to behavioral health providers by clarifying the options available to patients for treatment and the associated cost implications.