Skip to content

Suicide Prevention: Next Steps

By on

I’ve written here before about suicide prevention. I’m writing again about this incredibly important public health issue for several reasons. First and foremost, suicide rates continue to rise in the U.S.; this devastating scourge claims over 40,000 lives a year in our country. September marks National Suicide Prevention Month, a fact I want to underscore with this message. NIMH has a research agenda for suicide prevention, and we’re working to push harder on a plan of action that includes practice ready results. I wanted to tell you about that plan; in formulating it, we have focused on collaboration, in the spirit of our ongoing engagement with the National Action Alliance for Suicide Prevention. The National Action Alliance, a public-private partnership, has joined with the American Foundation for Suicide Prevention to focus on the goal of reducing the national suicide rate by 20 percent in 10 years. This ambitious but achievable goal is our inspiration as we consider how to capitalize on opportunities and plan NIMH’s research agenda.

What we know through collaboration

Collaboration has paid off even in the last few months, as we and our partners have learned more about how to identify those at risk for suicide and intervene in ways that can reduce that risk. Our past collaborations—with the Army and the Veterans Health Administration—have helped us develop screening tools and risk prediction algorithms. NIMH intramural and extramural research efforts have developed tools for implementation in real-world settings. Our current collaborations are testing the benefits of risk detection and pragmatic interventions. For example, findings from ED-SAFE, an NIMH-funded effort to test the effectiveness of brief screening for suicidality in emergency room settings, continues to make major contributions. Yes, emergency room-based screening, when applied universally, identifies more individuals at high risk than care as usual—a finding I wrote about last winter. But more remarkably, pairing this screening with a low-cost intervention—follow-up phone calls—results in significant decreases in subsequent suicide attempts over the next year. In a study coming out today, NIMH and extramural scientists collaborated on a mathematical modeling exercise that demonstrates that mail-, phone- and psychotherapy-based interventions could all be cost-effective if administered to patients identified as at-risk during emergency room visits. These data add to the armamentarium of evidence-based screening and intervention tools that we now know will help in our efforts to prevent suicide attempts and deaths.

Where our collaborations need to go next

There are two major next steps on NIMH’s plan of action to improve our practice-ready suicide prevention research. First, we must take what we already know works and learn how to implement it on a large scale. Second, we must focus further on gaps in our knowledge that would have a big impact in the near term if we could fill them in.

With respect to step one, implementation, we are working with the Substance Abuse and Mental Health Services Administration to coordinate Zero Suicide efforts. For example, we have two "practice ready" sets of tools that we know work well, at least in certain settings: risk detection algorithms, and risk reduction treatments. On the risk detection side, we have evidence from our work with the Army and the Veterans Health Administration that computerized algorithms, operating on administrative and clinical datasets, can identify those at high risk in these settings. We now are working together with a network of health care providers to learn how to implement these algorithms in general population settings. The eight general medical emergency room sites in ED-SAFE were capable of applying universal screening with good results. We need to collaborate with stakeholders, accrediting organizations such as the Joint Commission, and policy makers to increase uptake of these effective practices.

On the treatment side, we have reasonable interventions to offer that we know are cost-effective. A number of emergency room suicide prevention approaches were recently discussed at a May 2017 NIMH meeting focused on this topic. These include the post-emergency room interventions mentioned above. The collaborative care model that has been used to integrate behavioral health into primary care has been shown to be feasible for the treatment of multiple mental disorders across multiple health care settings, for diverse age groups, and it’s cost effective.1 This approach has reduced suicide ideation in depressed older adults in primary care,2 and it seems ripe for further suicide prevention testing, particularly with Medicare now supporting payments for psychiatric collaborative care—a decision that was built on years of NIMH research.3 We now must turn to our collaborators, both Federal and private, to help design and carry out implementation studies, which demonstrate feasibility for widespread adoption, as well as fidelity (how accurately are the treatments actually delivered) and effectiveness (how well do they do at reducing risk) in community settings.

Regarding step two, gaps in knowledge, answers to several open questions would have an outsized impact on our ability to prevent suicides. A major gap area is in our understanding of the time course of suicide risk: in those identified as at high risk, how long does this risk last? Can we follow the evolution of risk over the weeks and months after an individual is identified as at high risk? Are there indicators that would help us identify when that risk has abated?  Answering these questions would help us test and target interventions, perhaps even on the individual level. Systematic tracking of high-risk individuals would greatly assist in this regard. Another gap area follows from the mathematical modeling that predicts cost-effectiveness mentioned above; these models predict cost-effectiveness based on assumptions regarding how many people who visit emergency rooms are at high- or medium-risk of suicide; how many of them will participate in treatments; and how effective those treatments are. These models will become more useful as these assumptions are updated—informed by actual descriptive and intervention data that can come from multiple sources, including of course our collaborators as well as NIMH researchers.

Again, we will pursue a two-pronged approach. We need to study individuals prospectively to see when individuals are likely to be at risk, and which interventions work for whom, and for how long. But we need also to facilitate data sharing, and construction and use of registries and centralized data repositories (such as those at the Veterans Administration and our own at NIMH, as well as those at partners such as the Mental Health Research Network). These large datasets will enable fine-grained analysis of trajectories, a process which we are encouraging with specific requests for applications that call for secondary analyses to further our understanding of suicide risk and intervention benefits.

Together, our collaborative efforts to exploit what we already know and to further knowledge in these specific gap areas will help us achieve our collective goal of saving lives, and reversing the trend in suicide rates in the U.S.

References

1 Huffman JC et al. Essential articles on collaborative care models for the treatment of psychiatric disorders in medical settings: a publication by the academy of psychosomatic medicine research and evidence-based practice committee. Psychosomatics. 2014 Mar-Apr;55(2):109-22. doi: 10.1016/j.psym.2013.09.002. Epub 2013 Dec 25.

2 Bruce ML et al. Reducing suicidal ideation and depressive symptoms in depressed older primary care patients: a randomized controlled trial. JAMA. 2004;291(9):1081-91.

3 Press MJ et al. Medicare payment for behavioral health integration. N Engl J Med.  2017;376(5):405-407.