|Home | About | Journals | Submit | Contact Us | Français|
Health plans appear to be moving towards less stringent management, but it is not known whether behavioral health care arrangements mirror the overall trend. To improve access to and quality of behavioral health services, it is critical to track plans’ delivery of these services. This study examined plans’ behavioral health care arrangements and changes over time using a nationally representative health plan survey regarding alcohol, drug abuse and mental health services in 1999 (N=434, 92% response) and 2003 (N=368, 83% response). Findings indicate health plans' behavioral health service provision changed significantly since 1999, including a large increase in contracting with managed behavioral health care organizations. Some evidence of loosening administrative controls such as prior authorization implies easier access to services. However, increased prevalence of higher levels of cost-sharing suggests financial barriers have grown. These changes have important implications for enrollees seeking care and for providers working to meet patients' needs.
Managed care appears to be moving towards less stringent management,1,2 with enrollment shifting away from traditionally more tightly managed types such as HMOs towards preferred provider organizations (PPOs) and point of service (POS) plans. Yet, for one important type of care— behavioral health—it is not known how closely arrangements may be mirroring the overall trend.
Health plans frequently contract with managed behavioral healthcare organizations (MBHOs) to manage and deliver mental health and substance abuse services. Enrollment in any type of managed behavioral health care program was estimated at 164 million in 2002.3 Health plans and/or the MBHOs with which they contract have the potential to influence access to and utilization of behavioral health services in multiple ways. These organizations can use techniques aimed at providers by applying utilization management techniques, carefully selecting their networks, or using financial incentives; they can also focus on patient-oriented constraints and incentives such as benefit features.4 Essentially, there are direct financial versus other administrative mechanisms, and they can be directed at providers or patients. Many administrative mechanisms may have financial implications.
On the patient side, direct financial influences revolve around the benefit package, which specifies which services are covered, a prerequisite for access. The level of patient cost-sharing affects financial access. Meanwhile, administrative influences on patients include the way a health plan structures treatment entry, including the gate keeping process (for example, telephone call-in center versus primary care gate keeping) and utilization management mechanisms such as prior authorization. On the provider side, direct financial influences could include pay-for-performance approaches while administrative mechanisms such as paperwork required for utilization management purposes may also affect service delivery.
Both financial and administrative mechanisms can affect care. For example, utilization management techniques may result in utilization reductions through denials or approving less treatment than requested and/or through a “sentinel effect” in which the existence of the utilization management system itself may deter care despite typically low denial rates.5,6,7 Financial mechanisms have also been shown to affect behavioral health utilization, in prior studies on the effect of patient cost-sharing for outpatient visits8,9 and psychotropic medications,10 and in studies of provider response to changes in how they are paid.11,12 However, these effects of financial mechanisms may be amplified or attenuated by the administrative mechanisms used in managed care. For example, many have suggested that requiring health plans to raise behavioral benefits to ‘parity’ levels will no longer have the strong utilization impact that earlier research findings would imply, now that plans have other ways to restrain use besides benefit design.13 Thus, understanding how health plans provide behavioral health services requires attention to multiple domains.
Given the well-documented need to improve access to and quality of behavioral health services,14,15 and the large segment of the population covered by private health plans, it is critical to know about changes in how these organizations organize and deliver care. This study is the first to use nationally representative data to examine changes in health plans' behavioral health arrangements in detail across multiple domains during recent years, and provides the most recent data on this topic. Tracking changes from 1999 to 2003 is a particular opportunity as it allows for observation of behavioral health care evolution during a time of notable change in health care overall.2
A nationally representative survey of commercial health plans was conducted regarding alcohol, drug abuse and mental health services in 1999 and 2003. In 1999 434 health plans were surveyed in 60 market areas (92% response rate); in 2003 368 plans in the same market areas were surveyed (83% response rate). This telephone survey included an administrative module addressing behavioral health contracting, benefits, network management and provider payment, and a clinical module addressing primary care screening and treatment, entry into specialty care, utilization management, prescription drug formularies, quality improvement, and other clinically oriented topics. Each plan was asked about its top three commercial managed care products.
Typically, the executive director for the administrative module and medical director for the clinical module were able to respond to all topics in the survey, although some plans referred to the MBHO for some information. For some national or regional plans, respondents at the corporate headquarters level were interviewed regarding multiple sites. Items were asked at the product level within each market-area-specific plan.
The study is linked to the Community Tracking Study (CTS), a longitudinal study of health system change, funded by the Robert Wood Johnson Foundation.16 The primary sampling units were the 60 CTS market areas selected to be nationally representative. The second stage consisted of selecting plans within market areas. Plans serving multiple markets were defined as separate plans for the study and data were collected with reference to the specific market.
The 1999 sample frame was based on a CTS household survey that used responses to identify and survey insurers and health plans regarding plan characteristics. This survey yielded approximately 1,000 entities categorized as managed care plans. After excluding entities that were exclusively indemnity plans and those no longer present within market areas, 944 plans constituted the sample frame, from which 720 market-specific plans were selected with equal probability and without replacement across the sites. The plan sample was allocated to each market area based on the weighted estimate of plan enrollees in each site, with proportional distribution between PPO-only and HMO/multi-type plans. Of the 720 sampled plans, 247 were deemed ineligible because they had closed, merged or were otherwise unreachable, had fewer than 300 subscribers in the market, did not offer comprehensive health care products, served only Medicaid/Medicare, or for several other reasons. This left 473 eligible plans, of which 434 (92%) responded, regarding 787 eligible products. For the clinical module, 417 plans (88%) responded regarding 752 products.
For the 2003 survey, the 1999 sample of 720 plans was augmented with a national sample of plans not previously operating in the sites. Of the 110 new plans identified, 94 were selected, maintaining the same sampling rates as those in 1999. Thus, the 2003 sample frame consisted of 1054 plans, and the sample totaled 814 plans including all 1999 eligible sampled cases plus a sample of 1999 ineligibles and new plans. Of these 814 plans, 373 were ineligible, leaving 441 eligible plans, of which 368 (83%) responded and reported on 812 products. Of these, 347 plans (79%) completed the clinical module, reporting on 771 products. Four products that were "consumer-driven" (a new category in 2003) were excluded from analyses.
The following survey measures are included in this analysis:
Health maintenance organization (HMO), preferred provider organization (PPO), or point-of-service (POS).
Specialty (contracted with an MBHO for specialty behavioral health services; comprehensive (contracted with a single vendor for both general medical and behavioral health provider networks); and internal-only (all behavioral health services provided by plan employees or through a network of providers directly administered by the plan).
Whether a range of specific MH/SA services were covered under the most commonly purchased benefit package for each product.
For each covered service, whether prior authorization was required to initiate care.
What actions must enrollees and providers take for plan members to access regular outpatient specialty MH/SA care?
This section included organizational topics regarding whether the plan contracted with a pharmacy benefits management (PBM) firm or used a three-tier formulary (which typically have graduated co-payments for generic, preferred brand, and nonpreferred brand). Questions regarding newer (as of 2003) drugs and four brand-name antidepressants, five antipsychotics, and three medications for attention deficit hyperactivity disorder (ADHD) were also included.
Co-payment or coinsurance amounts for outpatient mental health, substance abuse and general medical visits for the most commonly purchased benefit package within each product.
The findings reported are national estimates. The data are weighted to be representative of health plans’ commercial managed care products in the continental U. S. Statistical analyses were implemented using SUDAAN software17 for accurate estimation of the sampling variance given the complex sampling design. When comparing results within a year by product type and contracting arrangement, pair wise t-tests were used to determine significance. Corrections were made for multiple comparisons (three pair wise tests for each comparison by product type or contracting arrangement) by using the Bonferroni correction. When conducting other bivariate analyses such as change analyses across years, t-tests or chi-square tests were used without correction.
In 1999 58% of products contracted with MBHOs (Table 1) and specialty contracting remained the most common arrangement in 2003 with increasing prevalence to 72%. This increase was driven by POS and PPO products moving from internal arrangements to specialty contracting. Specialty contracting remained above 80% among HMOs; however there was a small increase in the use of internal arrangements between 1999 and 2003.
Most MH/SA treatment services were offered by more than 90% of products in 2003 regardless of contracting arrangement (Table 2), although residential treatment was covered less frequently than other services. The picture was similar for 1999 (data not shown). Among all services, outpatient methadone maintenance was least likely to be covered.
Products with internal arrangements were less likely to cover some services than products with specialty contracting. Residential substance abuse rehabilitation services were covered by 52% of products with internal arrangements versus 94% of products with specialty contracting. Outpatient methadone maintenance was offered by 84% of products with comprehensive contracts, 68% of products with specialty contracts and 36% of products with internal arrangements. Residential mental health treatment was covered by 89% of products with specialty contracting, but only 42% of products with internal arrangements.
Prior authorization was required by 80% to 99% of products for inpatient hospitalization; residential treatment, rehabilitation and detoxification; and intensive outpatient MH/SA treatment, among products that covered the service (Table 3). Prior authorization was less common, but still usually required, for other services.
Prior authorization was most frequently employed by products with specialty contracting. For example, in 2003 63% of products with specialty contracts required prior authorization for mental health outpatient counseling, significantly greater than the 15% of comprehensive products that did so. Similarly, for outpatient detoxification in 2003, 61% of products with specialty contracts required prior authorization versus 41% for products with comprehensive contracts and 38% for products with internal arrangements.
Although prior authorization still predominates, there was a significant decrease in required prior authorization for outpatient mental health counseling, outpatient detoxification, and outpatient substance abuse rehabilitation, although there was a small increase in required prior authorization for both mental health and substance abuse intensive outpatient services. These changes were driven primarily by products with specialty contracting, including a reduction in prior authorization for mental health outpatient counseling from 94% to 63%.
Table 4 reports on requirements for mental health treatment entry; results for substance abuse were almost identical (data not shown). Products typically had no requirements for entry into specialty mental health treatment (41%), or required that either the health plan member or specialty mental health provider contact the plan, structuring this as a choice of how to meet the requirement (40%). Seven percent of products had the sole requirement that members contact the plan, while another 5% mandated solely that the provider must contact the plan.
There were significant differences by contracting arrangement. In 2003, products with comprehensive contracts were more likely (65%) to have no requirements for mental health treatment entry versus plans with specialty contracts (37%). Products with specialty contracts more frequently allowed either the member or provider to contact the plan (48% versus 29% of those with comprehensive contracts and 7% of those with internal arrangements). Nearly all products that required enrollees to call their phone center indicated that at least one outpatient visit was always authorized (data not shown). About 25% of plans with internal arrangements had some other type of requirements, with about one third each requiring PCP referral, allowing but not requiring PCP referral as one way to meet treatment entry requirements, or having a combination of mechanisms.
In 2003 54% of products contracted management of pharmacy benefits out to pharmacy benefit managers (PBM) (Table 5). Products that carved out behavioral health services to MBHOs were least likely to use PBMs. The vast majority of products used three-tier formularies to manage pharmacy benefits. This technique was used more frequently by products with specialty contracts.
The way that products used three-tier formularies varied significantly by contracting arrangement. Ten percent or fewer of products with specialty contracts covered all four of the newer antidepressants asked about on tiers 1 or 2, i.e., tiers with the lowest patient cost sharing. This was also true for the antipsychotic and ADHD medications studied. Plans with comprehensive contracting were most likely (70%) to cover all four antidepressants on tiers 1 or 2, while plans with internal arrangements were most likely to cover all five antipsychotics and all three ADHD medications on tiers 1 or 2 (46% and 56% of products, respectively).
The survey identified shifts in the cost-sharing structures employed by plans between 1999 and 2003 for MH/SA and general medical services. Table 6 includes results for mental health cost sharing; substance abuse results were almost identical (data not shown). Products reduced their use of coinsurance for behavioral health treatment and increased the use of co-payments. This is related to the increase in specialty contracting; MBHOs typically use co-payments for behavioral health visits, irrespective of whether the product is an HMO where co-payments are most common or a PPO where coinsurance predominates for general health care (data not shown). In contrast, coinsurance for general medical visits became more common, again reflecting the decline in HMOs.
High cost sharing has become more common since 1999 for both behavioral health and general medical services. The proportion of products requiring a higher level of cost sharing (defined as co-payments greater than $20 or coinsurance greater than 20%) increased from 26% to 42% for MH/SA and increased even more substantially for medical care from 4% to 41%. The difference in cost-sharing between medical and behavioral health care has shrunk, but this is due to more health plans falling into the higher cost sharing category for medical care in 2003.
Many studies have examined specific aspects of how health plans organize MH/SA treatment, but this is the first to undertake a detailed analysis of behavioral health services along multiple dimensions. The changes reported from 1999 through 2003 have implications for patients’ experiences in obtaining services and providers’ interactions with health plans and MBHOs on behalf of their patients. Often plans’ decisions about whether to contract with another entity for behavioral health services focus on considerations of costs, financial risk, and areas of core expertise.18 However, as discussed here, these decisions also have important implications for patients and providers.
The increased prevalence of specialty behavioral health contracting among plans indicates that MBHOs continue to play a large role in the commercial health sector. By 2003, carve-outs were predominant across all product types. However, recent reports of numerous health plans switching back to managing behavioral health benefits themselves after 200319 suggest that it will be important to track health plan contracting to see if the carve-out trend changes course on a national basis.
When enrollees have a choice among plans, those who choose HMOs expect that their choice of providers will be more constrained than in PPO or POS products. However, patients who select a PPO because they prefer to access a larger network of providers may find they have the same access to behavioral health services as those who selected an HMO. Regardless of the type of managed care product that they enroll in, most enrollees are likely to deal with an MBHO for behavioral healthcare.
From the perspective of primary care providers, the trend towards MBHOs could imply even greater fragmentation of patient care than is already typical as they are forced to deal with—or help their patients deal with— multiple entities. Although findings indicate that primary care gate keeping was rare in plans with specialty contracts, there may still be issues regarding care coordination across organizations and health plans. For specialist providers, however, the increasing trend toward carve-outs, coupled with the dominant market share of several large national MHBOs, may mean that there are fewer different organizations to deal with on behalf of their patients.
The large majority of products covered a wide range of behavioral health services, a clear prerequisite for gaining access to a full continuum of services. However, in the age of managed care, these nominal benefits do not guarantee access since strategies such as utilization management are designed to influence both initial access and total quantity of services delivered. These benefits may be viewed, therefore, as necessary but not sufficient. Required prior authorization is the norm for a range of behavioral health services, but less so in 2003 than in 1999. The reduction in prior authorization for regular outpatient care is noteworthy, since that is the level of care used by most enrollees accessing behavioral health services. This mirrors the move towards “less managed” managed care noted in general health care1,2. Presumably this would save providers as well as patients from the "hassle factor" associated with initial access. Other information that would be important in determining implications for access includes the quantity of services routinely authorized at one time, since there is evidence this can affect utilization6. Findings indicate that in 2003, the majority of products with prior authorization requirements initially authorized six or eight visits (data not shown). Information regarding exactly how the authorization process is conducted, and the perception of the level of ease and efficiency from the provider and patient points of view, would also be relevant although beyond the reach of these data.
Structural gate keeping requirements for regular outpatient behavioral health care are often minimal (although procedures may vary in length and detail required), primary care gate keeping is rare, and initial access is routinely approved. Ease of entering specialty treatment is important from the patient point of view, and also from the perspectives of both specialty and general medical providers whose attempts to guide patients toward appropriate resources would otherwise be hindered. However, significant issues can arise later in the process, the answers to which affect access to care in practice. First, actual availability of providers in the network and obtaining timely initial appointments can be an issue, particularly in some specialty areas such as child psychiatry. Second, the quantity of outpatient services allocated and the process for continuing treatment are important.
Understanding health plans’ organization of pharmacy benefits is crucial because of the large numbers of individuals treated with pharmacotherapy and the shift towards newer, costlier psychotropic medications.
Psychotropic drugs play an ever more important role in behavioral health treatment.20 By 2001, 75% of people treated for mental health or substance abuse received prescription drugs,21 and prescription drugs grew from 6 percent of total mental health spending in 1991 to 17 percent in 2001.22 Sales of newer antidepressants grew 19% between 1996 and 2001, versus only 4% for tricyclic antidepressants.22
In 2003, products with specialty contracts were least likely to have full availability of the newer medications studied to treat depression, psychotic disorders, or ADHD on lower cost-sharing tiers, making those medications more expensive for enrollees. It is important to recognize that for products with specialty contracts, the responsibility and financial risk for behavioral health services typically lies with the MBHO while the responsibility for prescription drugs remains with the general medical plan. Thus, MBHOs have an incentive to encourage the use of prescription drugs as a substitute for psychotherapy. One way to counteract this incentive is for health plans to make the patient cost-sharing for medications higher when there is a specialty carve-out, as seen in the results of this survey. However, much prescribing for depression and ADHD is done by primary care providers.23 From the patients’ perspective, therefore, when plans turn to increased cost-sharing to moderate incentives for prescribing by the MBHO’s network of specialty providers, this also has the effect of making patients’ out-of-pocket prescription costs even higher when primary care providers direct their care.
Cost sharing has been shown to influence the utilization of specialty behavioral health care.24,25,26 Understanding recent changes in behavioral health cost-sharing is complicated by the fact that there have been both shifts in the type of cost-sharing (moving from a percent of charges to fixed-dollar co-payments), and within each type, changes in the level of cost-sharing. These results suggest that financial barriers to access for behavioral health care have grown, as evidenced by the increased prevalence of higher levels of behavioral health cost sharing since 1999. One qualification is that these results assume constant fee levels, based on some anecdotal discussion suggesting that they were stagnant over the period in question.27 If in fact provider fees were increasing over time in some areas, enrollees in those areas would have benefited from the shift to co-payments, and the growth in burden would have been slower.
Comparing cost-sharing levels between 1999 and 2003 indicated the gap between mental health and general medical cost sharing had narrowed dramatically as the percent of products with higher cost sharing grew even more for general medical care than for behavioral healthcare. Thus, while the goal of parity in cost sharing between mental health and general medical care may not be closer, certainly proponents of this policy never intended that parity be achieved through increases in medical cost-sharing. From these results, it would appear that the closing gap between behavioral and medical cost sharing for outpatient care is, ironically, due to increases in both.
A general limitation of organizational surveys is the lack of information regarding how contractual arrangements and organizational requirements are operationalized in practice. For example, it is known that providers are required to contact the plan for prior authorization, but not how difficult it is to access the call center or how phone center staff interact with callers. Despite this lack of nuance, this study illustrates the broad landscape of health plans’ behavioral health service provision.
Both general medical and specialty providers treat commercially insured patients with mental health and substance use disorders. Unlike general medical care, however, behavioral health services often are organized differently because health plans may contract with MBHOs for mental health and substance abuse services. As the recent Institute of Medicine report on behavioral health care quality highlights, it is imperative to improve access to quality mental health and substance abuse services and to enhance integration between primary care and specialty treatment.28 The findings from this study inform efforts in this direction by providing a nationally representative picture of how these services are arranged currently, and the implications for access, including coordination of needed care.
The dominance of MBHO contracting in all types of managed care products underlines that the existence of a separate organization for specialty behavioral health care is the typical scenario to envision when considering behavioral health policy and practice relating to private health plans. In particular, this suggests the need to empirically investigate how this arrangement may affect coordination across specialty behavioral health and general medical sectors. The reduction in prior authorization requirements for outpatient care and the rarity of primary care gatekeeping point to an easing of “front-end” access that may be beneficial, but increased reliance on high cost sharing may constitute a barrier for some consumers. The dramatically closing gap in the percentage of products with high mental health versus general medical cost sharing points up an issue that pending federal parity bills cannot address: equal but burdensome cost sharing. However, it is also important to note that the impact of similar cost sharing may be even greater for behavioral health since demand for services has been found to be very sensitive to costs.24,25,26 The variation in extent of choice for new psychotropic medications by contracting arrangement also bears further investigation. Because patients often do not respond to the first psychotropic medication prescribed for common disorders like depression29, having a range of coverage medication options is critical. Placement of newer, brand-name drugs on higher tiers means that the drugs are available but more financial burden is placed on consumers. Both providers and policymakers should consider the potential impact of this approach.
These findings regarding private health plans' behavioral health service arrangements from 1999 to 2003 provide a nationally representative view of changes during this dynamic period in health care. Some of the changes found mirror overall health care trends such as the move towards less overt management in a context of greater patient cost-sharing, while other health plan features are unique to behavioral health (e.g., structuring access through behavioral health call centers). Given the rapidly occurring changes in health plans’ provision of behavioral health services, it will be important to track future evolution in this area.
The authors wish to acknowledge the contributions of Frank Potter and staff at Mathematica Policy Research, Inc. (survey design and statistical consultation, data collection), Richard Frank and Sharon Reif (comments on draft manuscript), Grant Ritter (statistical consultation), Melissa Morley, Maureen Stewart and Joanna Volpe-Vartanian (research assistance), Galina Zolutusky (statistical programming), Michele Hutcheon and Wanda Rifkin (manuscript preparation).
This study was supported by the National Institute of Drug Abuse grant #R01DA10915 and the National Institute on Alcohol Abuse and Alcoholism grant # R01AA10869. The 1999 round was also supported by the Substance Abuse and Mental Health Services Administration contract #98M-0028601.