Quality and Parity: New Payment Models for Behavioral Health | The Kennedy Forum

Quality and Parity: New Payment Models for Behavioral Health
September 8, 2017

Remarks of

Patrick J. Kennedy

Member of Congress, 1st District, Rhode Island, 1995-2011

CMS Innovation Center

Behavioral Health Payment and Care Delivery Innovation Summit

September 8, 2017

Click here to download the PDF version

Click here to watch the entire session.

“The Kennedy Forum’s leadership helped the Centers for Medicare and Medicaid Services (CMS) implement the collaborative care model which includes outcomes measures for our National stakeholders.”

Patrick Conway, MD, MSc

CMS Deputy Administrator for Innovation and Quality


Thank you for the invitation to provide an overview of urgently needed payment models that have been developed to improve mental health and addiction treatment in this country. During my 16 years in Congress, and with the Kennedy Forum over the last four years, I have advocated holding health insurers accountable by requiring parity of coverage for mental health and addiction treatment. I have also encouraged holding treatment providers accountable by promoting the adoption of evidence-based measures. I’m pleased to have an opportunity today to talk about initiatives that can achieve these goals.

Payment systems today often are not designed to incentivize high-quality, evidence-based behavioral health care, promote integrated care for Americans suffering from both mental health and physical conditions, or pay for the most promising and innovative service delivery methods. Most of us have at least some familiarity with models such as collaborative care, coordinated specialty care for those experiencing first-episode psychosis (FEP), and telemedicine. These treatments have been proven to work, but they are not used enough. So we need to make sure that payment systems pay for them in a way that incentivizes uptake.

We must also recognize that while it is important to talk about the specific solutions identified below in terms of payment reform, we cannot live in denial about the magnitude of the concurrent opioid and suicide epidemics our country faces. Recently we saw the devastation wrought by Hurricane Harvey in Texas while the most powerful hurricane ever seen in the Atlantic Ocean—Irma—imperils the East Coast. However, the destruction from these horrific natural disasters pales in comparison to the devastation of the opioid and suicide epidemics both in terms of lives and dollars lost. And, these national disasters threaten those in coastal metropolises and those in rural areas alike. Every day our emergency rooms and morgues experience a Hurricane Harvey four times over because of drug overdoses and suicides. If that is not considered and declared a national emergency subject to immediate Executive and Congressional attention, then our priorities are misaligned. It took Congress nearly two years to set aside $1billion to address the opioid epidemic in the 21st Century Cures Act, yet it only took the House two days to authorize nearly $8 billion for Hurricane Harvey relief.

Promoting Behavioral Health Parity

Before we talk about new payment systems, we must address parity implementation or the lack thereof. There cannot be true behavioral health payment system reform until we fully enforce the Federal Parity Law.  This means addressing non-quantitative treatment limitations such as medical management techniques that may result in denials of medically necessary care, and “experimental” exclusions of proven treatments.

We must also increase reimbursement for behavioral health providers, which results in inadequate networks and has contributed to a workforce shortage. And, when care is delivered using capitated or bundled payments, we must improve the risk adjustment for mental health and substance use disorder (MH/SUD) to reduce the incentives to engage in behavior that limits access to MH/SUD care. Basing reimbursement rates on cost data collected before full implementation of parity doesn’t make sense, and recreates the disparities of the past.

Capitated and bundled payments can shortchange mental health and addiction treatment if not accompanied by real quality metrics. When we measure the “success” of treatment in models that incorporate value-based payment (VBP), we need to use outcome measures that actually tell us about the effectiveness of treatment such as symptom severity, reduced morbidity and mortality, and ability to function at home, work, and in the community. Other metrics, like number of visits, length of stay, or spending per episode of care may be useful to monitor, but they do not show that treatment works.

Funding Alternative Payment Models

We must develop systems that incentivize when appropriate, and require when necessary, the most evidence-based and cost-effective interventions for mental health and addiction treatment.  Recently, the Thomas Scattergood Behavioral Health Foundation, along with The Kennedy Forum and the Peg’s Foundation, funded an issue brief highlighting several innovative Alternative Payment Models (APMs). The paper is titled Payment Reform and Opportunities for Behavioral Health: Alternative Payment Model Examples. This paper incorporates ideas from some of the leading individuals and organizations working on payment reform throughout the country and many of the key recommendations are incorporated below.  We are recommending APM covering: 1) preventative services; 2) early interventions; 3) screening and initial treatment in primary care; 4) SUD specialty care; 5) integrated care; 6) measurement-based care; 7) supportive services and other non-health care spending; 8) implementing new technologies; and 9) hospital and community partnerships.

Preventative services: Many indicators can identify children at elevated risk for developing mental health conditions and substance use disorders, such as high Adverse Childhood Experience (ACE) scores. We also know that the social determinants of health in a child’s environment are critically important for the overall mental health and wellbeing of the child as she progresses towards adolescence and adulthood. APMs must link children and families to the necessary social services in conjunction with treatment services.

One model that does this, Accountable Communities for Health at Nemours Children’s Health System, is a place-based collaboration among healthcare and social sectors in which partners are held financially accountable and jointly responsible for meeting predetermined metrics and goals. Building on the Innovation Center’s Accountable Health Community Model Track 3, the Accountable Communities for Health for Children and Families would test the use of an integrator organization aligning clinical and community efforts to address family risk and protective factors and connect families with needed social services to optimize the health of the family and child.

Early intervention: We must fund treatment for those experiencing first-episode psychosis (FEP) or prodromal symptoms. The “RAISE” protocol provides a tested blueprint for successful and cost-effective interventions for persons with FEP with its coordinated specialty care (CSC) for individuals experiencing FEP. We need to direct payments to services identified in RAISE by incentivizing outreach and engagement, treatment, and rigorous follow-up services.

An ideal APM model for CSC programs would use a case rate or bundled rate that covers the full array of services and supports delivered in evidence-based CSC programs. In developing APMs for CSC programs, case and/or bundled rates should be available for the first two years of intensive treatment, with adjusted rates that support young people in maintaining gains as they transition to less intensive follow-up care. APMs should support continuity of care and ongoing positive treatment outcomes for individuals experiencing psychosis associated with schizophrenia.

Screening and initial treatment in primary care: We must ensure the early identification and treatment of MH/SUD conditions by PCPs and other non-behavioral health providers (NPs, PAs, etc.). This must include screening for a multitude of conditions, not just depression. Providers should be paid to screen their patients for MH/SUD conditions, provide initial treatment, and make appropriate referrals to behavioral health specialists. And, patient outcomes must be rigorously tracked and analyzed.

We recommend a demonstration project with one or more ACOs in which they screen for all common mental and substance use disorders. In addition, the ACO would track outcomes for all these conditions by using a quantifiable and standardized symptom rating scale. The ACO would carry some risk for these outcomes and could also use a variety of other APMs as part of the project.

SUD specialty care: We must expand Medication Assisted Treatment (MAT) by using innovative and evidence-based alternative payment models. The Patient-Centered Opioid Addiction Treatment Payment (P-COAT) is an alternative payment model designed to improve outcomes and reduce spending for opioid addiction. The model creates a bundled payment structure representing three phases of care. Bundles can be used in collaboration or separately depending on patient needs.

The Patient Assessment and Treatment Planning payment is a one-time payment to support evaluation, diagnosis, and treatment planning. The “initiation” of MAT service bundle also is a one-time payment but covers services related to the initiation of outpatient MAT, including supervised induction of buprenorphine therapy, appropriate psychological or counseling therapy, and care management and coordination services. The “maintenance” of MAT service bundle is a monthly payment covering all care related to ongoing medication, psychological treatment, and coordination of social services necessary to remain in treatment following initiation.

We also should encourage private sector initiatives to promote value-based SUD treatment. Health plans and managed behavioral health organizations such as Beacon Health Options and Optum have been rolling out bundled MAT payment mechanisms that cover a variety of services. Health plans should make available outcomes data from these programs, and ensure that the programs are compliant with the Federal Parity Law.

Another example of incentivizing opioid maintenance treatment would be to add a billing modifier if a person who has been diagnosed with an opioid use disorder: 1) is being treated by a doctor, counselor, or in an inpatient or outpatient mental health or substance abuse treatment facility; and 2) is on maintenance medication with one of the three FDA approved medication.  In such cases, CMS could pay a 25% additional payment to the current baseline for those billable services. The use of temporary medications for detoxification would be excluded; we need to incentivize the provision of maintenance medication treatment consistent with the medical evidence.

And, individuals must remain on buprenorphine while incarcerated. Many fatal overdoses occur when people are released into the community with reduced tolerance after suspension of MAT while in jail.

Integrated care: We must promote effective approaches like the Collaborative Care Model, which has been shown to be effective in dozens of studies. Thankfully, Medicare Part B now includes billing codes approved for use in 2017 for reimbursement of services using the Collaborative Care Model (CoCM). Those codes are:

  • G0502 – this allows payment for CoCM for the first 70 minutes of the first calendar month for initial psychiatric collaborative care management of behavioral health care manager activities
  • G0503 – this allows payment for CoCM for subsequent psychiatric collaborative care management, which includes the first 60 minutes in a subsequent month of behavioral health care manager activities

G0504 – this allows payment for additional CoCM for any month, which can be used either with the initial or subsequent psychiatric collaborative care management and covers 30 additional minutes in a month of behavioral health care manager activities.

A separate code, G0507, has been established for general behavioral health integration services as well. This code is used for care management services for behavioral health conditions and covers at least 20 minutes of clinical staff time directed by a physician or other qualified health care professional per calendar month. CMS plans to convert these codes for CoCM services into CPT codes in 2018.

In addition, CMS is proposing to allow Federally Qualified Health Centers (FQHCs) and Rural Health Clinics (RHCs) to receive separate payments for CoCM and Behavioral Health Integration (BHI) services in 2018. This would allow CoCM and BHI services to be encompassed in the definition of “primary care services” that lead to assignment of beneficiaries to a Medicare ACO (accountable care organization), and may encourage ACOs to adopt the CoCM.  CMS should also incentivize the use of peer support in conjunction with CoCM.

However, we need to take some additional steps to better reimburse uptake of Collaborative Care Codes by increasing prospective payment & other incentive models which would increase utilization of G Codes. There are multiple ways to achieve this.

One method would be through a multipayer system. BHI is an emerging priority for private payers and state Medicaid agencies, and is best supported through aligned incentives from multiple payers. Consequently, the BHI model could include public and commercial payers that desire to align with the model, while also complementing state-based initiatives, where available and feasible.

Another important innovation would be the creation of a “consultation” fee. When primary care initiates the consultative relationship with a psychiatric professional, the consultant could receive a supplemental G-code for providing decision supports and ad hoc eConsults for a patient population under their shared care, as well as any telepsychiatry payments.

And, of course, an appealing development would be the use of performance-based incentive payments. Primary care practices and psychiatric consultants would jointly receive a prospectively paid performance-based incentive payment for all patients receiving CoCM services. They would share or pay back the payment, depending on performance on key metrics tied to CoCM: (1) quality measures such as a) behavioral health electronic clinical quality measures (eCQMs) (e.g., Depression Remission and Response) measured at the practice-level, b) patient-reported outcome measures,  c) patient experience surveys (CAHPS); and (2) utilization measures that drive costs such as emergency department and in-patient utilization.

One model that is currently in place that we recommend is a bundled payment approach developed in Minnesota. Among the first to create a multi-payer system for collaborative care, this project was spearheaded by a unique partnership that included the state’s six largest commercial health plans, the Minnesota Department of Human Services, and medical provider groups in the state. Together, these groups and organizations agreed that improving depression care was a priority and that the FFS reimbursement system available at the time was inadequate for primary care practices to support effective depression care management.

Under DIAMOND (Depression Improvement Across Minnesota, Offering a New Direction), primary care providers implemented collaborative care for depression and could bill for a negotiated bundled monthly payment rate, which was deigned to cover associated clinical costs. Honored by all the major commercial insurance companies in the state, the availability of this bundled payment mechanism was enough for about 80 diverse primary care practices to accept the burden of collaborative care startup costs (such as hiring care managers and registry development), because of the promise of at least breaking even if they enrolled enough patients from different payers in their collaborative care program.

Measurement-based care: We must further develop and require the use of measurement-based care and link this to payment and incentives. Without true quality measures, ACOs and other systems with value-based payment or bundled payments will continue to skimp on providing high-quality behavioral health care. Most ACOs do not screen for or track non-process outcomes measures for other common mental health or substance use disorders.

This lack of ACO quality measurements incentivizing evidenced-based behavioral healthcare is a missed opportunity for payers, health systems, and patients. Given demonstrated higher healthcare costs and reduced work productivity associated with patients who have mental and/or substance use disorders, it is a logical choice for ACO contracts to couple behavioral health financial risk with corresponding robust and specific performance markers.

Private health insurers also should encourage the implementation of true measurement-based care. For example, Aetna has incorporated behavioral health outcomes metrics in value-based contracts (VBCs) with dozens of providers, including use of evidence-based care, access to psychiatry and other services, and continuity and collaboration of treatment. Outcomes data from these types of programs should be made publicly available, and we must ensure that the programs are compliant with the Federal Parity Law.

Supportive services and other non-health care spending: We must encourage the reimbursement for supported housing, life skills/job training, transportation, environmental modifications and other support services.  This can be addressed within bundled payments that feature braided funding with dollars coming from other social service funding from federal, state, and county sources.

For example, Vermont’s Support and Services at Home (SASH) program provides community-based support services and affordable housing communities using a combination of funding from Medicaid, Medicare, the Department of Vermont Health Access, the Department of Health, and the Department of Aging.

Implementing new technologies: We must incentivize innovations such as telemedicine, which can improve access to and coordination of care. The Veterans Health Administration (VHA) has been a national leader in telemedicine for years. This is partly due to the multiple unique characteristics that make the VHA a well-suited organization for telemedicine: a uniform electronic medical record, a lack of state-specific provider licensure regulation, and a relatively tech-savvy patient subpopulation (from recent experience serving in the military). This APM facilitates the expansion of telehealth services by incentivizing the use of existing payment codes.

The Kennedy Forum also has identified the other leading technological advances that will move the field into the future in our recent issue brief, New Technologies for Improving Behavioral Health. These include: predictive modeling health information, active and passive assessments, remote peer support, biofeedback and many others.

Hospital and community partnerships: Another improvement that is badly needed is improved coordination among hospitals and community providers. This would offer the opportunity to increase efficiency and reduce unnecessary hospital admissions. Currently, one in eight visits to an emergency department (ED) in the US involve MH/SUD. More than half of inpatient MH/SUD stays begin in an ED. This method of treating individuals with behavioral health conditions is extremely costly and does not deliver high-quality care.

An intervention targeting patients with psychiatric conditions in the ED could have the potential to address their BH needs sooner and improve the likelihood of reducing health care and social service related costs. A new model could use Behavioral Health Coordinators (BHC) to enhance care coordination for patients with a primary or underlying behavioral health (BH) condition in partnership with a community BH provider and ED. Under this model, patients arrive in an ED, are triaged, and receive medical care from ED staff as is customarily the practice. Patients are then referred to the a community BH provider if they meet eligibility criteria, e.g., repeated admissions.

Payment incentives would be used to target patients with psychiatric conditions in the ED using the initial G0502 code for initial psychiatric collaborative management billing code. Participating BH community organizations and hospitals could also receive a capitated dollar amount per member per month (PMPM) for each beneficiary enrolled in their program based on risk and intensity of services needed.

Final Thoughts

Brain fitness must be treated with the same respect and dignity as any other organ of the human body.  Much needs to be done to create incentives for insurance companies, health systems, providers and others to treat each individual holistically.  A primary goal is to create a level playing field in terms of how both mental health and physical care are paid for.  CMS can assume a leading role in the U.S. by adopting many of these alternative payment model recommendations referenced in my testimony today.  When CMS leads with these type of payment reforms, the rest of the industry will follow.  Now is the time to make a difference and promote meaningful reform.  By funding these types of value-based care programs, everyone will benefit.  Our population health goals cannot be achieved without the right payment incentives to promote evidence-based, behavioral health interventions.

To read an Issue Brief entitled Payment Reform and Opportunities for Behavioral Health: Alternative Payment Model Examples which includes policy recommendations to improve quality and reduce the cost of U.S. health care, click here.

The Kennedy Forum

I founded The Kennedy Forum in 2013 as a way to convene cutting-edge thinkers who are united by the potential for reform in behavioral health service delivery made possible by new laws, revolutionary technologies and enhanced understanding of effective services and treatments.  Our inaugural event in October of that year called for The Forum to develop a platform to advance thinking across a host of issues in our field.  The Kennedy Forum is organized to drive real, lasting, meaningful policy change to fulfill our vision.

Today, The Kennedy Forum’s work is not singular in its focus. We are promoting mental health coverage through a series of initiatives, which include:

  • Ensuring health plan accountability and compliance with the letter and spirit of the parity law by educating consumers, providers, and regulators, so that each group holds themselves and others accountable for proper enforcement.
  • Establishing ways to promote provider accountability through evidence-based outcomes measures that are validated and quantifiable.
  • Implementing proven collaborative practice models that promote the integration of mental health and substance use disorder services into mainstream health care.
  • Using technology to optimize electronic/digital communications and enhance assessment/treatment tools.
  • Promoting brain fitness and wellness, which includes identifying opportunities to translate neuroscience research findings into preventive and treatment interventions.