Redesigning Federal Health Insurance Policies to Promote Children’s Mental Health

Effective interventions to enhance child and adolescent mental health can be focused on rehabilitation, treatment of acute conditions, prevention of new conditions, or promotion of mental well-being. Although a growing body of science suggests that the most cost-effective and moral solutions focus on early promotion and prevention, the U.S. federal agencies that spend the most on child and adolescent mental health continue to focus their payments on rehabilitative and treatment services. Specifically, the federal Medicaid and Children’s Health Insurance Program (CHIP) programs are the dominant revenue sources for behavioral health services to children and adolescents in the U.S. Although there are other funding streams that cover early education and developmental services for children (e.g., Maternal and Child Health Services Block Grant [Title V, Social Security Act]; Head Start early intervention services [via the Individuals with Disabilities Education Act, Part C]); Child Care Development Block Grant; and the Maternal, Infant, and Early Childhood Home Visiting Program), Medicaid and CHIP cover almost half of all children with special healthcare needs. Medicaid and its Early Prevention, Screening, Diagnosis and Treatment Program (EPSDT) and CHIP offer health and behavioral healthcare coverage for children whose parents are unable to afford private health insurance. Although these programs are designed nationally, they are administered at the state level.

Medicaid alone covers a total of 70 million individuals, including a large share of vulnerable populations—32% of children (44.2 million) and 46% of pregnant women—and has an annual cost of $475 billion to federal and state taxpayers.1 However, because Medicaid’s EPSDT and CHIP emerged from adult-oriented insurance models, they are primarily oriented toward providing services when problems arise, rather than supporting the conditions for children’s healthy development. Access to most services under Medicaid requires that a licensed provider, in an approved or licensed site, assign a diagnosis for an illness to an individual—meaning that to access behavioral health services, a mental health problem has to have already developed.

Children, however, are not small adults. Neurodevelopmental studies, animal models, and studies of brain plasticity in infancy have established how important early childhood is for forming capacities for lifelong development,2,3,4 marking this as a period of both great vulnerability and great opportunity.5 The neural connections that form the building blocks of brain development begin before birth and continue into early adulthood. Therefore, early interventions can have profound and long-lasting effects, and a wide range of interventions addressing neuro-regulation and social relationships are effective.6,7,8 Just under half of U.S. children experience at least one adverse childhood event (ACE), and one in ten experiences three or more; these negative events place them at high risk for later negative health and mental health outcomes. Although a strong body of evidence traces the impact of ACEs on lifelong development,9,10,11 an equally strong body of evidence exists about effective prevention and early intervention services to mitigate these negative outcomes.12

Therefore, the opportunity is to redesign Medicaid, EPSDT, and CHIP so that they optimize children’s development and functioning and promote the capacity for lifelong health and quality of life. We recognize that the changes we suggest will not affect all U.S. children, because these policies cover only a portion of the nation’s child population—those in families with lower incomes. Nevertheless, making these changes would set a precedent. We also recognize that decisions about modifications are made at both federal and state levels and that there is considerable latitude for states to adjust coverage. In this paper, we focus primarily on policy levers where federal guidance to states will promote coverage likely to optimize children’s healthy development. A thoughtful redesign will attend to the unique developmental issues of children, including the systems in which they live, learn, and seek services (i.e., home, school, neighborhood, and primary healthcare). We argue for a redesign to explicitly address four issues with greater relevance to children’s development: periodicity (i.e., the developmental continuum, requiring ongoing assessment of functioning); provider capacity (requiring expansion of service providers to include skilled nonprofessional providers and increased training); place or setting expansion to offer continuous, coordinated, responsive, and flexible coverage across a range of locations; and payment reform to ensure that services are adequately reimbursed in different contexts and to promote investments in prevention and early intervention programming. We conclude with specific recommendations.

An orienting feature that distinguishes children from adults is their developmental plasticity. To promote and optimize children’s functioning, ongoing assessment and monitoring of core functional markers is needed to guide the rapid deployment of interventions—as conditions develop—for children with high-risk profiles (e.g., three or more ACEs). Therefore, measurement and monitoring of children’s functioning over time must be frequent (periodic) and coordinated across systems and must take into account developmental change, proxy reports, and temporal stability of measures as complicating factors in measurement psychometrics.

What Is Missing?
The current design of Medicaid’s EPSDT and CHIP falls short. The Center for Medicaid and Medicaid Services (CMS) 2013 policy guidance13 on the assessment of childhood functioning recommends using the periodicity schedules set by Bright Futures and standardized assessments (e.g., Ages and Stages Questionnaire and Pediatric Symptoms Checklist) in general medical settings for children. These are necessary but insufficient steps. These nonspecific recommendations are implemented poorly in most settings. First, state EPSDT programs have noted consistent shortcomings in the actual delivery of developmental and behavioral screening and treatment in primary care and general medical settings. Second, Medicaid and CHIP reimbursement policies exclude payment for behavioral health services provided in numerous other settings that are not licensed or certified as specialty sites, where high-risk children are often found, such as homes, general classrooms, juvenile justice settings, foster care homes, and youth development programs. Third, current measurement standards are not tied to key child functioning goals, such as kindergarten readiness, third-grade literacy, or high school graduation. This is because pediatric measures for state reporting are focused narrowly on follow-up after hospitalization or diagnosis-specific services (e.g., for ADHD), not on functioning. The implication for providers and payers: that the process of developmental and behavioral screening is more important than outcome.

There are substantial workforce shortages and a likely misallocation of providers who care for children’s behavioral health. This includes shortages of pediatricians, case managers, special education teachers, social workers, psychiatrists, and behavioral healthcare workers, particularly in rural areas.14,15,16,17,18 Coupled with these workforce shortages are issues around lack of training—especially in developmental assessment and psychosocial interventions—for general healthcare providers, teachers, and other frontline professionals who work with children. Training deficits are most pronounced in knowledge about basic developmental competencies, recognition of behavioral health issues, and evidence about effective early interventions.19,20 There are too few specialists, and the generalists do not understand the emerging science on mental health promotion, prevention, and early intervention that could alter the trajectory of many children.

What Is Missing?
Policy guidance is absent to support widespread workforce training on the use of effective prevention and early intervention programs. Also absent is guidance promoting expansion of the behavioral health workforce to include family peer support, youth peer support, and telemedicine. These are low-cost, acceptable, and effective alternative services. Medicaid guidance on funding sources for family and youth peer support services and telemental health for children is inconsistent and poorly understood or applied. The most recent federal guidance on peer support21 provides only general guidelines on scope of and access to peer support services. Some states are using Medicaid administrative matches to pay for these services. Similarly, federal guidance continues to support low reimbursement rates for telemedicine, compared with reimbursement for face-to-face services.

Promoting children’s behavioral health and resilience is most effective when services can be provided in places where young children develop that include nontraditional healthcare settings, such as homes, schools, and other natural contexts (e.g., childcare settings and faith-based settings). This is because many effective interventions focus on improving skills of parents and other primary caregivers (e.g., early education staff). It also means that providers (i.e., case managers, early learning specialists, and healthcare specialists) need to be trained in how to provide services within these nontraditional settings. However, both Medicaid’s EPSDT and CHIP favor delivery of services in more traditional settings (i.e., outpatient and hospital-based settings).

What Is Missing?
The provision of services in nontraditional mental health service settings, such as in the home and in childcare settings, is covered under Medicaid’s Home and Community-Based Services Waiver program, which also includes the provision of intensive in-home services, but these services are largely case management for already identified “patients,” and coverage in schools is limited. The covered services do not currently include prevention or promotion services.

Paying for children’s prevention and health promotion services requires thinking differently about return on investment, the balance of long-term savings vs. short term gains, and the availability of appropriate codes with which to support these services (i.e., nonmedical necessity).

What Is Missing?
There is no code enabling reimbursement for monitoring or tracking of pediatric referrals to behavioral health services, thus making it unlikely that follow-through will be tracked and measured and that children will receive follow-up care. In addition, the current design incentivizes short-term contracts with managed care organizations, thus making long-term investments, which could include prevention and early intervention programs, difficult to secure. The payment models that exist are largely based on fee-for-service configurations. To address children’s developmental functioning, however, and to ensure that no child is missed, population-based payments linked to full accountability need to be integrated into policies. This could include pediatric accountable care organizations (ACOs) and value-based purchasing mechanisms, discussed below in recommendations.

There is no code enabling reimbursement for monitoring or tracking of pediatric referrals to behavioral health services, thus making it unlikely that follow-through will be tracked and measured and that children will receive follow-up care.

An important step is to develop measures that allow for frequent monitoring of children’s development and functioning—measures that capture childhood’s incremental changes, with explicit attention to modifiable community and neighborhood factors22,23,24 that have an impact on children’s health and development. Although CMS’s 2016 overhaul of Medicaid managed care25 requires states to collect a standard set of encounter data, these data do not include periodic measurement of children’s development. States are required to report on Healthcare Effectiveness Data and Information Set (HEDIS) measures, but measures for children are focused largely on hospitalization rates or on specific diagnoses (e.g., ADHD and depression). The National Academy of Science, Engineering, and Medicine (NASEM) recommends an alternative approach: the Vital Signs: Core Metrics for Health and Health Care Progress report identifies 15 domains for capturing a range of health outcomes; however, most of these domains are focused on adults, not children.26 A recent initiative, organized by NASEM’s Forum on Promoting Children’s Cognitive, Affective, and Behavioral Health, is identifying developmentally responsive functional measures for children to parallel or complement the broader Vital Signs work. The pediatric measures will capture cross-system issues, especially important for children; account for continuity of care and social determinants of health that comprise children’s environments; and include community and neighborhood measures.22,23,27

There are limited payment options under current mainstream federal policies for services delivered by professionals other than licensed medical or health providers. Federal guidance is needed to clarify the competencies—not professional degrees—needed by a wide range of professionals (e.g., early childhood paraprofessionals and parent and youth peer support providers) who have the potential to provide evidence-based, appropriate early intervention and prevention services. For example, only about a third of states currently use Medicaid funding for peer parent support services,28 although evidence for these programs, as well as a national certification process and pool of certified candidates, exists. And, although 34 states report that Medicaid covers an “early childhood mental health specialist,” no state covers services that aim to benefit an entire classroom or group of children through consultation or group training activities. Opportunities also exist to expand provider capacity by increasing federal reimbursement for telemedicine behavioral health coverage and for states to expand their coverage of this service, which reaches rural areas where such services are critically needed; currently, the Medicaid telemedicine provider reimbursement rates are far below both Medicare and commercial insurance rates.29

Recent federal efforts to increase workforce capacity via evidence-based training and technical assistance do not cover early childhood prevention and intervention programming or providers. They need to do so. The Substance Abuse and Mental Health Services Administration’s recent program announcements for regional Mental Health Technology Transfer Centers30 and Prevention Mental Health Technology Transfer Centers31 represent a significant development. However, the announcements focus on “those with serious mental illness” and “provide training and technical assistance services to the substance abuse prevention field.” Neither of these efforts is likely to provide training and technical assistance to help establish early intervention programming needed for children and parents to remediate the effect of ACEs and prevent mental illnesses from developing. This is a missed opportunity.

Some states are enacting statewide training initiatives to retool their current workforce, including New York State’s Evidence-Based Treatment and Dissemination Center and Minnesota’s Evidence-Based Practices initiative, both of which utilize the Managing and Adapting Practice system.32 States are also addressing workforce shortages by creating credentialing processes to train and certify peer support providers, including parents and youth.33,34,35,36,37 Family peer advocates (FPAs) are a group of providers whose work improves family engagement, access, and knowledge about quality.38,39,40,41,42,43,44 A national training and certification program also exists for peer parent support providers in children’s mental health.45 New York State has made a significant investment in family peer support, training and certifying nearly 800 FPAs, and is expanding its training and certification efforts by using different models (e.g., train-the-trainer),46 including launching online training modules,47 in anticipation of the growing use of FPAs now that Medicaid reimburses for those services. Federal guidance is needed to help states expand reimbursable services by using nontraditional workforces, including family and youth peer support specialists.
Telemedicine and digital services also counteract the effects of workforce shortages.

Coverage of services needs to be expanded to include services delivered where children spend most of their time (e.g., homes, schools, and communities). This should include support for evidence-based parenting programs, early intervention skills training, and school readiness (e.g., early literacy and reading). These kinds of services do not fit within the traditional model because they are delivered outside of specialty mental health clinics and do no necessitate a mental health diagnosis. NASEM’s recently formed Collaborative on Healthy Parenting in Primary Care has identified specific policy opportunities for delivering evidence-based parent support programs and family-focused interventions (e.g., Triple P and Incredible Years) in primary care,48 but these are not reimbursable under Medicaid or CHIP. In addition, the U.S. Preventive Services Task Force recently recommended both public and private insurance coverage for family-focused interventions as a preventive measure.

Embedding reimbursement codes for these kinds of services would facilitate children’s healthy development. Current Medicaid regulations do not overtly promote parent enrollment in evidence-based parent training programs; only 12 states report that Medicaid pays for parenting programs (and only two of these states, Michigan and Oregon, require providers to use an evidenced-based parenting program); 37 states (76%) report that they do not cover these services,49 despite evidence of their effectiveness6,50,51 as well as their significant cost-benefit savings.52 Finally, current Medicaid policy design sometimes promotes discontinuous coverage and disruption of services for many of the highest-risk children when there are minor changes in residence, foster care transitions, and eligibility changes due to disability.

Another potential avenue for expanding place-based coverage is through the Home and Community-Based Services (HCBS) Waiver program, authorized by the Social Security Act. This program allows for the provision of standard medical and nonmedical services in the home or community. The target of HCBS services is individuals with established diagnoses and usually those who already have tangible impairments. As such, this “flexible” model is not able to provide developmental interventions for children who are high risk of not flourishing (i.e., who have multiple ACEs), because these children are not yet diagnosed or disabled. HCBS services have begun to revolutionize care for children with autism spectrum disorders, with covered behavioral interventions aimed at improving functioning. Such an approach would be very useful for children who have experienced multiple ACEs. This would require redefining eligibility, perhaps by limiting access to children with high ACE scores (e.g., three or more), linking the services to full-service pediatric practices, and defining the scope of intervention (e.g., number and type of sessions). Federal guidance could consider piloting, via a limited waiver opportunity paired with a robust evaluation, HCBS services coordinated by full-service pediatric practices (using collaborative care bundled payments) for young children at high risk of developing a behavioral health diagnosis. The services should be focused on evidence-based interventions and circumscribed as appropriate in terms of amount, scope, and duration.

More sophisticated, complex, creative, and sustainable financing mechanisms are needed to fund early prevention and intervention programming. Some states are finding innovative ways, in the absence of federal guidance, to invest in the early years. Oregon’s “At-Risk Codes” allow payments for early intervention to reduce the development of serious mental health conditions later,53 and New York State’s “First 1,000 Days of Medicaid” recognizes that a child’s first three years are the most crucial years of development.54

New Payment Models
Several CMS efforts have attempted to address the development of new payment models, including CMS’s 2017 Request for Information on Pediatric Alternative Payment Model Opportunities,55 which sought input on the design of alternative payment models to improve the health of children and youth covered by Medicaid and CHIP and focused on concepts that encourage pediatric providers to collaborate with health-related social service providers at the state and local levels. However, no further follow-up or request for proposals (RFP) has been issued (as of July 2018). CMS also funded 31 demonstration projects under the Accountable Health Communities Model and developed a Health-Related Social Needs Screening Tool.56 Opportunities exist to provide additional technical assistance to communities to further develop these needed models, with a focus on pediatric ACOs and expanding value-based purchasing mechanisms to include pediatric health and behavioral health. To date (July 2018), no RFPs or further federal policy regulations or policy guidance exist to support broader implementation of these alternative financing models/mechanisms.

Pediatric ACOs
Pediatric ACOs began nearly two decades ago. However, utilization of this model is not widespread and is exclusively affiliated with large hospital systems. As such, the variability across models—due to state-to-state differences in Medicaid and lack of federal guidance on structure and financing—has made existing pediatric ACO models nationwide difficult to compare, align, or regulate for best practices.57 Two recent evaluations of pediatric ACOs, one of early adopters and another profiling hospital-based models, found a need for more complex financial models to account for the long-term investment and short-term savings. Further, the cost-savings goals tend to take precedence over measurement of quality, and there is a need for development of ACO quality measures.57,58

One example of a pediatric ACO model is Nationwide Children’s Hospital Partners for Kids (PFK), one of the oldest pediatric ACOs (founded in 1994) and a full-risk, population-based model that receives a Medicaid per-member per-month age- and sex-adjusted payment. PFK has demonstrated lower costs than fee-for-service or managed care organizations. Importantly, this slowing in cost growth was achieved without losses in overall quality or outcomes of care.59 In addition, PFK’s care coordination reduced inpatient and emergency department utilization.60

Value-based purchasing
In the past decade, many states have undertaken significant efforts to overhaul their Medicaid plans,61,62 with reforms aimed at moving away from fee-for-service toward increasing the value of the care they are providing. However, these efforts are happening in the absence of federal guidance, leading to uncertainty and inefficiencies. These include integrating primary and behavioral healthcare and replacing fee-for-service payments with reimbursements based on clients’ attainment of health outcomes.63,64 Two-thirds of children and youth covered by Medicaid and CHIP are now in managed care.65
Examples of state innovations include Oregon, which has created coordinated care organizations that are locally governed to address community needs on a single global budget, and New York State, which is implementing the Delivery System Reform Incentive Payment program and Performing Provider Systems to restructure Medicaid to reduce avoidable hospital use by 25% over five years. However, few children are hospitalized, and this focus in New York effectively excludes pediatric services.66 Minnesota is piloting Certified Community Behavioral Health Centers and Health Home initiatives to coordinate care across settings and providers.67 These state innovations are using more complex financial models to try to provide more flexible services, but they are not primarily targeted at children or at promoting children’s healthy development. This needs to change.

Return on Investment tools
States would also benefit from the development of tools that allow them to calculate their return on investment (ROI) in early prevention and intervention programming. A recently published Commonwealth Fund “return-on-investment calculator” (Assessing Risks and Rewards of Integrating Social Services with Health Care) enables policymakers to weigh the value of partnerships/investments in social services that are likely to improve medical outcomes.68 An adaptation of this tool for investments in early prevention and intervention programming for children could provide state policymakers with data regarding their investments in these prevention and early intervention programs. An important caveat, however, is that it is unethical to hinge treatment on purely financial calculations. ROI tools for prevention are not required for other health conditions (e.g., cancer and heart disease). Nevertheless, it would be useful for planning purposes to know probable ROIs for prevention programs for children. ROI analyses can be a useful tool to encourage expanded coverage by states. For example, since the late 1990s, the Washington State legislature has directed the Washington State Institute for Public Policy (WSIPP) to calculate the ROI to taxpayers from a variety of education, prevention, and intervention programs and policies. Today, WSIPP provides ROI data for an extensive range of programs and policies to guide decision making at the state level.52

Mainstream federal policies that address children’s health needs—Medicaid, including EPSDT, and CHIP—are not designed to promote healthy development or to address the unique needs of children. We call for a redesign of these programs to include development of billing codes and payment options to enable frequent monitoring of children’s development (periodicity); training, technical assistance, and coverage of a range of credentialed, nontraditional providers; expansion of covered settings; and adoption of payment models at a systems level to create a sustainable structure. Supporting children necessitates equal attention to supporting—even nurturing—their families and communities to create the conditions for positive development and life-long functioning. Federal health policy must keep pace with the scientific evidence on what is known to create the conditions for future generations of children to thrive.