Preventing suicide is an important public health priority, but community-based programs are often not based on evidence that their strategies can work. Many suicide prevention programs end up wasting money, as well as opportunities to save lives.
In this article, I take a quick look at a recent review of brief interventions to prevent suicide after a person has made a visit to an emergency or urgent care service for suicidal ideation or attempted suicide or has been discharged from an inpatient stay for those reasons.
Such persons are at high risk of subsequently dying by suicide, especially if they break contact with mental health care, which they often do. Yet emergency departments are not typically equipped to provide follow up care. They depend on patients engaging with mental health care elsewhere.
The authors were interested in interventions that could be delivered in a single contact, perhaps with brief messages afterward. The interventions were intended to engage patients with mental health care and reduce the likelihood of a subsequent suicide attempt.
The article is paywalled, but if you click on the link, you can at least get a look at a detailed abstract.
Doupnik SK, et al. Association of Suicide Prevention Interventions with Subsequent Suicide Attempts, Linkage to Follow-up Care, and Depression Symptoms for Acute Care Settings: A Systematic Review and Meta-analysis. JAMA Psychiatry. 2020;19146. doi:10.1001/jamapsychiatry.2020.1586
The authors undertook a systematic search of multiple databases for titles and abstracts in the past 20 years that included relevant key words. This search yielded almost 4000 articles, but ultimately only 14 studies were considered eligible for inclusion in the review, with outcomes for 4270 patients.
Each article included at least one of these outcomes: subsequent suicide attempts, linkage to mental care, and depression symptoms at follow-up.
Components of the interventions included brief contact in the form of telephone calls, handwritten notes, and text messages. Three studies included coordination of care, which involved back and forth communication between the clinic team referring the patient and the team receiving the patient for follow-up outpatient care. Five studies involved a Safety Planning Intervention:
(1) identifying personalized warning signs for an impending suicidal crisis, (2) determining internal coping strategies that distract from suicidal thoughts and urges, (3) identifying family, friends, and social places that can distract from suicidal thoughts and urges, (4) identifying individuals who can help provide support during a suicidal crisis, (5) listing mental health professionals and urgent care services to contact during a suicidal crisis, and (6) lethal means counseling for making the environment safer.
The review reports the pooled effects of the brief intervention were a moderate increase in engagement in mental health care. The authors estimated that these results represented a 22.5% increase in receipt of follow-up care in patients receiving the intervention over usual care.
There was a small reduction in subsequent suicide attempts, estimated to be a 3.5% reduction or 78 fewer suicide attempts in the 2241 patients receiving the intervention.
However, there was no significant reduction in depressive symptoms associated with patients receiving the intervention versus being in the control condition.
I gave a look at the figures and supplementary tables and found that the studies included in the meta-analysis were generally small, except two larger projects reported in three articles. Neither was a randomized trial, but both involved a matched comparison, either between sites receiving the intervention to similar sites providing usual care or the same sites before and after the intervention. These two projects accounted had three-quarters of the weight in the analyses of subsequent suicide attempts.
The authors did careful analyses and found that removal of any one study did not affect their results. However, removing the results of the two projects eliminated the significance of the reduction in suicide attempts associated with receiving the intervention.
I took a further look at the large project that provided the two effect sizes. It was an impressive study of patients receiving the Safety Planning Intervention (SPI+) after being recruited from the emergency departments of US Veterans Health Administration hospitals. Results from the larger component of the study were:
The SPI+ was associated with about 50% fewer suicidal behaviors over a 6-month follow-up and more than double the odds of engaging in outpatient behavioral health care.
Veterans have about a fifth higher risk of suicide than the general population, but the VA hospitals have better coordination of care than other parts of the thoroughly fragmented American health care system.
Lower-income patients receiving Medicare have more coordinated care in the public health system and fewer worries about insurance coverage. However, persons receiving insurance through their employers may face copayments and those who are unemployed may lack insurance coverage altogether.
Bottom line is that SPI+ intervention may be less effective in the mostly less organized aspects of American health care.
The authors of the review stated
“We were not able to examine whether brief suicide interventions ultimately reduced suicide deaths because most studies in the review did not include death as an outcome.”
But isn’t reducing deaths the purpose of suicide prevention? I doubt a meta-analysis with a pool of only 4270 patients is large enough to register a significant effect on deaths, even in a very high-risk group recruited from EDs and discharges from inpatient stays. At that level of risk, most patients will still not die by suicide, regardless of whether they get an intervention.
I have been an external scientific advisor for two decades of multi-region, multilevel community-based programs in Europe and a significant decrease in deaths has eluded us. Even when these projects relied on a composite outcome of deaths plus serious efforts at self-harm, we have mostly just missed an effect, pooling outcomes from five countries or more.
I was delighted to discover a few years ago the first study demonstrating a significant reduction in deaths among teens at high risk of suicide.
The intervention involved assigned them an adult buddy to give them support and keep them in treatment after they were discharged from an inpatient stay. The study only achieved a modest effect when the teens were followed up to 11 years.
The study actually found no significant decrease in suicide. Only when the investigators combined deaths due to suicide with deaths due to substance and alcohol use but judged not to be intentional self-harm. Apparently, teens at high risk of suicide as often die from using drugs and alcohol as from a deliberate attempt to end their lives.
Like the current review, this study was only a signal, not a definitive conclusion about how we can prevent deaths by suicide.
Every death by suicide, every effort at serious self-harm is a tragedy and a loss, not only for a person but for a circle of people for whom that person’s life has meaning.
Looking back, family members and professionals can usually identify some point where something could have been done differently that might have saved a life. But predicting the past does not allow us to predict the future and precisely when and where the next suicide will occur and what could be done to prevent it.
Every suicide is a tragedy, but it is inevitable that somewhere the next suicide will occur. Believing otherwise only leads to the needless blame of family members and punishing professionals who only in hindsight can identify what could have been done.
Suicide is infrequent enough and has such diverse causes that no single risk factor or combination will predict accurately what specific individual will die by suicide. Most individuals who die by suicide were not at high risk. Preventing suicide is complicated.
I sometimes advise members of the Canadian government on how to spend scarce resources wisely in efforts to deal with suicide as a major public health issue. I recently reminded them of what they were up against.
In 2017, the last year for which full data are available, there were 4157 suicides in Canada. The Canadian population in 2017 was 36.5 million, but this population is spread across 3.85 million mi.² 72% of the population is concentrated within 150 km (93 miles) of the nation’s southern border with the United States, but that is one of the longest borders the world. Additionally, there is great diversity within the Canadian population and that diversity matters when suicide prevention is being considered.
What then is to be done? We know that restriction of the means to prevent suicide is an evidence-based strategy that will ultimately reduce death by suicide on a population basis.
We know how ready access to affordable primary care and specialty mental health care reduces a risk factor for suicide and has its own benefits beyond reducing suicide. How this is achieved will differ with different populations in different contexts.
The current well-done review with its conservative analyses and interpretations has some modest but important messages. Brief interventions may well be able to keep high-risk persons in contact with mental health services and just may reduce acts of intentional self-harm. Those are excellent goals in themselves.
Special thanks to Dr. Stephanie Doupnick for her answers to my questions about her team’s review.