This Commentary describes limitations of mental health parity requirements in ensuring access to insurance coverage for mental health treatment and surveys regulatory options employed by states in Medicaid managed care programs as supplements to parity that can further reduce the risk of inappropriate denials of coverage. Part I describes inappropriate barriers to coverage for mental health treatment, Part II describes parity as a solution to such barriers, Part III explains that parity is not enough, and Part IV surveys supplements to parity in Medicaid managed care.
I. Inappropriate Barriers to Coverage for Mental Health Treatment
Mental illnesses, including substance use disorder, routinely go untreated or undertreated,1 with tragic results.Reference Tovino2 Barriers to insurance coverage for evidencebased treatment are one important reason for this mental health treatment gap. Even those who are enrolled in a health insurance plan face barriers to treatment in the form of coverage exclusions (which prohibit coverage for certain treatments and services), network limitations (which limit coverage to care from certain providers), prior authorization and step therapy requirements (which require insurer approval in advance of coverage or require patients try and fail low-cost therapies before receiving coverage for costlier, evidence-based therapies), and utilization review (which sees insurers denying coverage on a treatmentby- treatment basis for lack of “medical necessity”).3
There is reason to believe that such barriers are often inappropriate, i.e., not justified by the need to ensure quality or to avoid waste. Insurers’ economic incentives reward them for denying care to the sick or imposing coverage barriers that reduce administrative costs even while sacrificing quality or creating waste.Reference Lawrence4 Moreover, reports have identified examples of insurers imposing barriers on coverage that are not in patients’ best interest.5
II. Parity as Solution
Scholars and policymakers addressing the problem of inappropriate barriers to insurance coverage for mental illness have largely focused on federal and state parity laws as a solution.Reference Heller6 Parity “refers to the financing of mental health care on the same basis as the financing of physical health care.”Reference Tovino7 For example, the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity Act of 2008 prohibits insurers from discriminating against mental illness vis a vis other illnesses in designing and administering their plans.8 A beneficiary subject to unlawful discrimination may bring an administrative complaint or file a lawsuit, and state and federal regulators also have authority to bring actions identifying and seeking to penalize parity violations.
Several excellent recent articles have documented insurers’ continuing imposition of inappropriate barriers to insurance coverage for mental illness despite these parity laws.9 These articles have called for more vigorous enforcement of existing parity laws as well as legislative and administrative expansions of parity laws to close identified gaps. This focus on enforcing and bolstering existing parity laws is appropriate and important. If the objective is to ensure that people with mental illness have access to evidence-based treatment, however, parity is not enough.
There is reason to believe that such barriers are often inappropriate, i.e., not justified by the need to ensure quality or to avoid waste. Insurers’ economic incentives reward them for denying care to the sick or imposing coverage barriers that reduce administrative costs even while sacrificing quality or creating waste. Moreover, reports have identified examples of insurers imposing barriers on coverage that are not in patients’ best interest.
III. Parity is Not Enough
Parity is not enough for three reasons. First, as recent parity-focused articles have acknowledged, parity laws suffer from significant underenforcement. Greater enforcement would certainly help. This underenforcement problem is stubborn, however, because the underlying undertreatment problem is structural. Insurers impose inappropriate obstacles on coverage for mental health because of stigma, because of high costs, and because of their general economic incentive to limit coverage and reduce administrative costs.Reference Lawrence10 Each of these factors shapes insurers’ underlying incentives in designing and administering plans.
As Fineman explains speaking of traditional nondiscrimination laws generally, “equality, [] reduced to a prohibition against certain targeted discrimination, has proven an inadequate tool to resist or upset persistent forms of subordination and domination.”Reference Fineman11 Identification and correction of wrongs ex post is an inherently incomplete response to structural causes of discrimination such as insurers’ perverse incentives. These incentives influence every decision an insurer makes; by contrast, non-discrimination law impacts decisionmaking only insofar as the entity anticipates that particular acts of discrimination will be identified and penalized. Yet, as McGuire and Baker each explain, many decisions made by insurers in administering plans are not readily susceptible to documentation or monitoring.Reference McGuire12
Second and relatedly, parity's incompleteness itself may be inequitable. The subset of wrongful barriers to coverage corrected by parity laws is not random. Because the enforcement mechanism relies on complainant- or regulatorinitiated proceedings, parity laws most strongly affect that subset of barriers that beneficiaries choose to challenge by filing a complaint or bringing an action. In other contexts, legal scholars have noted ways that hinging enforcement on complainant identification biases enforcement in favor of better-connected, better- educated, and wealthier victims with the cultural capital and trust in the legal system necessary to identify and take action to rectify wrongdoing.Reference Galanter13 Moreover, in the context of managed care, scholars have found particular challenges to navigating coverage for enrollees with severe mental illness.Reference Hall, LaPierre and Kurth14
Third, even perfect parity between coverage for mental health treatment and coverage of other illnesses would still not be enough. Conflating perfect parity with adequate coverage assumes that insurers do not create inappropriate barriers to coverage for “physical” illnesses. But that is not true. Insurers have incentives due to adverse selection to impose wrongful barriers on coverage for all varieties of illness, and insurers often inappropriately limit coverage.Reference Lawrence15 This creates a gap between what Frank calls “parity in principle” and “parity in law.”Reference Frank16 Especially in the midst of public health crises associated with untreated mental illness, including substance use disorder, the aspiration of our law should not be that inappropriate barriers to treatment for mental illness are no larger than inappropriate barriers to treatment for less stigmatized illnesses. The aspiration of our law should be the elimination of all inappropriate barriers to mental health treatment!
IV. Supplements to Parity in Medicaid Managed Care
Medicaid managed care is a fertile context in which to consider legal tools that can supplement or complement parity in further reducing inappropriate barriers to coverage for mental illness.17 Although by default Medicaid is a government-run program, seventy-five percent of Medicaid's 72 million enrollees are in a managed care plan because their state has opted to privatize the program at least in part.18 A state that converts from a government-run health insurance system to a privately-run system is thereby creating a structural problem — insurers’ perverse incentives to deny needed care — that is a significant source of barriers to care.Reference Jacobi19 This structural problem can be mitigated by careful incorporation of legal correctives.
States employ a range of tools beyond parity to prevent or correct inappropriate barriers to behavioral health services. These tools are mutually complementary and have differing strengths and weaknesses.
A. Benefit Mandates
A clear-cut way to reduce artificial barriers to coverage is to require coverage for particular treatments or services, or, relatedly, to prohibit cost-sharing or prior authorization.20 Benefit mandates carry two interrelated drawbacks. First, it is usually impossible to mandate that a treatment or service be covered in all cases, because such a mandate creates the risk that Medicaid would be forced to pay for medically unnecessary care; this would be costly and potentially harm patients. As a result, state mandates may forbid coverage exclusions— policies prohibiting coverage for certain treatments and services—but leave insurers to consider case by case the medical necessity of such treatments and services. Second, mandates cannot police insurers’ exercise of discretion where they inevitably have it on necessarily particularized questions like individual medical necessity determinations.
B. Risk Adjustment
At least twenty-five states employ risk adjustment to some degree in their Medicaid managed care programs.Reference Juhnke21 Risk adjustment systems modify payments to insurers based on factors that predict the health or sickness of patients, such that insurers receive more money for covering expectedly-sicker individuals, and less for covering expectedly-healthier individuals. One purpose of risk adjustment is to mitigate insurers’ underlying economic incentive to discriminate against the sick (and thereby push costlier-to-insure enrollees out of coverage).Reference Buck22 In other words, rather than accept that insurers will seek to discriminate and try to police such discrimination (as do parity laws), risk adjustment seeks to reprogram insurers to mitigate their incentive to discriminate against the sick in the first place.Reference Lawrence23
Risk adjustment carries particular promise because it is capable of actively incentivizing insurers to facilitate (rather than impede) treatment for mental illness. Scholars have proposed that policymakers recognize that the status quo involves pervasive undertreatment, and so “risk adjust for the system they want, not the system they’ve got.”Reference Layton, McGuire and van Kleef24 This could be done by recalibrating risk adjustment payment models to calculate payments based on efficient costs (what costs would be in a world in which patients received coverage for all appropriate treatments), not actual costs (incurred in our current world in which patients often go without appropriate treatment). Doing so would encourage insurers actively to seek to bring patients in need of mental health treatment (or treatment for other targeted illnesses) into the risk pool, rather than avoiding (or “lemon-dropping”) such individuals. Scholars have also proposed drawing from the risk adjustment system in the Netherlands, which includes a separate, more-inclusive risk adjustment formula for mental health care that incorporates social factors “such as poverty level and household composition”.Reference McGuire, van Kleef and van Veen25
Risk adjustment is no panacea due to current technological limitations, which leave data collection and assessment accuracy limited. Thus, at best risk adjustment merely mitigates insurers’ perverse incentives to avoid sicker patients, it is not presently capable of eliminating them. Moreover, risk adjustment creates the potential for fraud, waste, and abuse as well as operational challenges that can further frustrate its effectiveness.
C. Insurance Coverage Appeals
State Medicaid programs generally permit enrollees to appeal a claim denied on medical necessity grounds, which may include denied mental health or addiction treatment.Reference Worthy26 Such appeals feature biased, internal review at early stages of appeal but ultimately lead to an independent, external review for those claimants persistent enough to pursue multiple layers of appeal.27 It is not uncommon in such systems to see low win rates at early stages of appeal coupled with high win rates at later, external stages, but vanishingly small numbers of claims persisting to that level.Reference Vukadin28
Appeals are an invaluable recourse for those claimants able to make use of them, but they suffer from some of the same limitations as parity. A beneficiary must, in the midst of illness, recognize their right to appeal and have the wherewithal to appeal through multiple stages in order to obtain relief. This leaves appeals capable of mitigating but not eliminating the problem of inappropriate denials and creates a problematic inequity in the population protected.
D. Reporting Requirements
Available data is insufficient to judge the extent of inappropriate barriers to coverage for mental health treatment in Medicaid, let alone the relative usefulness of the various tools discussed here. When a state transitions control of its Medicaid program to a private managed care plan, it also transitions all of the data generated in operating that program to that private entity. States seek to level the informational playing field by imposing reporting requirements on their managed care contractors to varying degrees.29 As a recent example, in May of 2019 New Jersey passed a law in response to perceived shortcomings of parity that requires, among other things, annual reporting by insurers of carefully-delineated metrics regarding their administration of mental health coverage.30
The diversity of Medicaid programs is today proving a barrier to the understanding that is an essential first step to assessing the extent of inappropriate obstacles to coverage and effectiveness of various regulatory tools in addressing such obstacles. As the Medicare Payment Advisory Commission put it: “[D]ata submitted by managed care plans to states and by states to CMS vary in their consistency, availability, and timeliness. This variability creates challenges for analyzing and monitoring managed care programs and limits the ability to compare states.”31 A mechanism to harmonize state data collection and permit states to work together to understand and control their Medicaid managed care programs could be a first step toward meaningful reduction in the barriers to coverage associated with state adoption of Medicaid managed care.
E. Behavioral Health Carve Out
Finally, a state that has confidence in its public administration systems may be unwilling to trust mental health coverage to managed care given the limitations of existing tools for preventing managed care plans from creating inappropriate barriers to coverage. Pennsylvania, for example, is one of several states in which mental health is carved out of the Medicaid managed care plan that oversees most treatments and services; instead behavioral health is managed by government actors in single county authorities. This choice creates a challenge of fragmentation of behavioral and physical health services, however, which has led advocates to note the potential care coordination benefits of an integrated approach.32
Pennsylvania is as of this writing considering legislation to end its carve-out and transfer responsibility for behavioral health to a Medicaid managed care entity.33 This legislation would require that the resulting managed care contract “[e]nsures that each recipient receives high quality, comprehensive health care services in the recipient's local area[.]”34 As this commentary has explained, in light of insurers’ underlying incentives to create obstacles to mental health treatment, a carefully-considered combination of tools is essential in any such contract for it to hope to comply with such a mandate. Parity alone is not enough.
Conclusion
Health policy in the United States has made significant strides toward promoting access to treatment for mental illness, in large part thanks to statutory parity requirements. But there is a long way to go, and we will need to use tools beyond parity to get there. In the Medicaid managed care context, states have employed various tools beyond parity to promote access to treatment for mental illness including benefit mandates, risk adjustment, insurance coverage appeals, reporting requirements, and behavioral health carve-outs. Careful assessment, development, and implementation of such tools is essential in addressing undertreatment for behavioral health issues, including mental illness and substance use disorder.
