ARM’s packed program requires attendees to make difficult choices about which sessions and presentations to attend, and this year proved to be especially challenging. As a researcher focused on behavioral health, I was pleased to see that mental health and substance use was the most common track this year, with over 195 abstracts (153 of which were posters) and nine sessions. In my view, this reflects the enormity of the health, social, and economic burden of mental illness and substance use, as well as the opportunities we have in effecting policy change in this space.
To organize this summary of selected presentations, I am borrowing Jonathan Purtle’s framework on policy in implementation science, which he presented during a Mental Health and Substance Use Interest Group organized session at ARM 2023. According to this framework, policy can fall into 1 of 3 main buckets: (1) policymaking; (2) policies; and (3) policy implementation.
Policymaking
In policymaking, the focus is on process and the motivations (i.e., evidence) for policy change. A few qualitative studies fit this category. Samantha J. Harris’s work on New Mexico’s 2022 law eliminating cost-sharing for behavioral health found that there is optimism around the new statute, and that it can benefit individuals with complex conditions the most. In another study by Jennifer Hefner and colleagues, community members from Columbus, Ohio expressed not just their frustration with biased policing but also the need for mental health training for officers who respond to mental health and substance abuse crises.
Several descriptive studies also highlighted gaps in policy and practice that can lead to policy change. Stephanie Ann Romero’s research, for example, revealed that only 28 percent of youth-serving mental health facilities offer LGBTQ-specific services, despite the higher risk for suicide and mental health issues faced by sexual and gender minority youth. Youngran Kim presented evidence around the seasonality of suicidal ideation and suicide attempts among adolescents, which implies that enhanced interventions from schools, clinicians, and public health systems are needed during critical periods. Finally, separate presentations by Aryn Phillips, Andrea Acevedo, and Eden Bernstein, among others, showed how one-size-fits-all approaches to dealing with alcohol use disorder (AUD) fails to recognize the diversity of AUD patients and their needs.
Policies
In Dr. Purtle’s framework, policy can be a strategy or the fixed context in which implementation occurs. Much of the impressive evaluative work presented at ARM fall into this category, and showed us which policies work and, more importantly, which do not. Ezra Golberstein and colleagues showed that elimination of cost-sharing for mental health and substance abuse in New Mexico lead to an increase in dispensed prescriptions and a reduction in out-of-pocket spending. Angie Waliski presented on her team’s award-winning paper demonstrating the feasibility and acceptability of an emergency department (ED)-based, peer-delivered safety planning for individuals at risk of suicide. This work showed that a policy that supports peer-delivered services (a strategy) can work in the ED where clinicians are willing to implement suicide prevention interventions (the context).
Advanced analytic methods were also used to demonstrate the effectiveness of different policies. Huiri Dong used interrupted time series analysis to analyze the effect of state-mandated naloxone co-prescribing with opioids, which the Centers for Disease Control and Prevention has recommended since 2016. They found mixed immediate increases in co-prescribing across states, but also declining growth rates over time, likely suggesting that these policies are limited. Soham Sinha and colleagues used machine learning methods to show that paid employment is a strong predictor of successful opioid use disorder (OUD) treatment among justice-involved individuals. Using mathematical modeling, Phillip Hughes showed that giving qualified psychologists prescriptive authority may reduce the unmet need for mental healthcare, with some variability between states.
Policy implementation
In policy implementation, barriers and facilitators to successful execution or scale up are evaluated. For instance, Matthew Goldman analyzed the 2022 rollout of the 988 National Suicide Prevention Lifeline and found that call volumes increased in all states, but earmarked state funding for call centers is only sufficient in about half of the states. To enhance responsible prescribing of opioids post-surgery, Zachary Ward and his team demonstrated that simple e-mail reminders or “nudges” to surgeons reduced above-guideline opioid prescriptions.
Several presentations also showed efforts to improve ongoing implementation projects. Justin Knox and colleagues developed an “observational-implementation hybrid approach” that incorporates implementation science principles in an observational study among Black sexual minority men to improve the delivery of evidence-based practices. Jessica Chen and her team administered a survey to understand attitudes of primary care and behavioral health staff in providing OUD care and found that new strategies could help alleviate the burden of OUD care delivery experienced by the participants. Similarly, qualitative research by Sophie Roe and others showed that clinical stakeholders including clinicians, administrators, providers, and care coordinators had favorable view of their OUD treatment program but found resources for psychosocial care to be lacking.
ARM presenters showed us creative ways in which policy can increase access to care, rollback ineffective programs, and improve the implementation environment. I was impressed by the depth of the work I’ve reviewed here, as well as the hundreds other pioneering abstracts not mentioned here. I can’t wait to see all the exciting work in behavioral health in Baltimore next year!