Some Alternative Payment Models Are Improving Behavioral Health

Alternative payment models (APMs) are taking steps to improve behavioral health, but the models need more to make large strides, according to a study from the University of Washington and Harvard Medical School.

The study from published in JAMA Network Open last week detailed the findings of a literature review on 17 APM implementations in mental health and substance use disorder (SUD) care.

The review found that APMs, which ranged from simple infrastructure payments to capitated reimbursements, were associated with improvements in process-of-care outcomes, reductions in mental health and SUD utilization, and decreases in spending.

However, information on whether the alternative payment models were effective at improving mental health and SUD clinical outcomes was largely missing from studies and other evaluations of the models.

Researchers also found evidence of gaming and adverse selection among participants of some alternative payment models.

“Our findings highlight important gaps in the literature on APMs in MH/SUD care,” researchers stated. “[A]lthough process-of-care measures are expected to be associated with clinical outcomes and are commonly used in APM evaluations for their feasibility, the association between these 2 outcome types is not always certain, often leaving the associations between APMs and clinical outcomes unclear.”

APMs are designed to incent the delivery of high-quality, cost-effective care, making them ripe for use in the behavioral health care space.

Mental health and substance abuse disorders have been associated with total healthcare spending that is two to three times higher than the national average, researchers reported. The conditions also commonly co-occur.

Through incentives to deliver more efficient, comprehensive, and team-based care, APMs have the potential to reduce behavioral health care spending while improving outcomes for patients oftentimes managing several conditions at once.

However, the healthcare industry has yet to capitalize on the opportunity in contemporary APM implementations, researchers said.

The study found that about half (9 of 17) of the APMs analyzed targeted SUD populations, while four models targeted mental health populations, and the rest targeted a combination of mental health and SUD, broadly defined.

Additionally, most APMs (11 or 65 percent) focused on adults, while just two models (12 percent) focused on children or adolescents.

Improvements to the APM evaluation process could help to boost APM implementation.

Researchers first recommended that future APM investigations apply rigorous design and analysis through either a randomized clinical trial or natural experience that includes robust observational design methods and sensitivity analyses, such as those used in category 3A of the Oxford Centre for Evidence-Based Medicine framework.

They also advised future APM investigations to include gaming and adverse selection evaluations, as well as at least one clinical outcome in addition to process-of-care measures that incent delivery of coordinated care and other evidence-based behavioral health care.

Clinical outcomes were only used in about 30 percent of the APM evaluations analyzed, while process-of-care measures predominated in the evaluations, with nearly 90 percent of APMs evaluated on their associations with these measures.

APMs were also more likely to be evaluated for their association with utilization (65 percent) and spending (53 percent) more often than clinical outcome changes.

But getting clinical outcome measures relevant to high-quality, cost-effective mental health and SUD care has been a challenge for APM developers.

“The majority of the SUD measures that are currently recognized are process and structural measures, rather than outcome measures — and among these outcome measures, few address the high-risk acute and chronic conditions specific to people living with SUDs,” the Center for Health Care Strategies, Inc. pointed out in a 2018 paper.

Challenges specific to the development of outcome measures for behavioral health APMs stem from the ongoing collection of evidence on the appropriate treatment of SUDs and the inconsistent collection of relevant data, the group stated.

Additionally, longstanding reluctance to share unnecessary patient outcomes due to compliance concerns has contributed to quality measurement gaps in the field.

But progress is being made with behavioral health APMs. CMS recently announced the participants in an initiative that aimed at developing alternative payment models that improve prevention, early identification, and treatment of behavioral and physical health needs in children covered by Medicaid.

Groups including the American Society of Addiction Medicine, American Medical Association, and yhe Alliance for Addiction Payment Reform have also released APMs focused on mental health and SUD care.

APMs are constantly evolving as providers uncover the best way to deliver high-quality, cost-effective care (and measure it), and industry leaders are setting their sights on the behavioral health care space as an area to expand value-based care.