How Suicide Research Is Harming Patients

Regardless of well-funded research, initiatives, commissions, foundations, mental health first aid, telephone numbers, or other tools distributed for safer suicide care, from 2000 to 2016, the suicide rate in the United States increased among most age groups from 10.43 to 13.92 per 100,000 (CDC, Vital Signs, 2017). As dreadful as these statistics are, between 1999 and 2017, the National Institutes of Mental Health (NIMH) reported that over 100,000 U.S. youth, ages eight to 24, committed suicide, and 1,309 were children ages five to 12.

Recently, the Suicide Prevention Resource Center (SPRC) and the American Foundation for Suicide Prevention (AFSP), the nation’s two largest nonprofits dedicated to saving lives, reported astonishingly conflicting data—without a shared or deliberative public explanation.

According to SPRC Director Elly Stout, “About 95 percent of all persons who commit or attempt suicide have a diagnosed mental disorder (December 2018).” Earlier in the year, AFSP reported, “Recent data show that the majority of those who die by suicide don’t have a mental health diagnosis (CDC, Vital Signs, June 2018).”

These two mutually exclusive announcements cannot be true in the same context at the same time. The probability intersection is empty and meaningless. Yet, “95 percent” and “majority” are not necessarily collectively exhaustible, and their union may not cover all unconventional desperate events in the remaining statistical suicidal space, i.e., autonomous, irresistible suicide without forethought (Copelan, 2006).

Nevertheless, these staggering statements, coupled with a 50-year peak in U.S. suicide deaths, not only attest to the fact that mental health researchers and clinicians are woefully unprepared to differentially diagnose suicidal emergencies, but also frames a disturbing binary fallacy: that there are only two explanations for the currently devastating suicide dilemma.

Is this either/or thinking a tactic? An ingrained tendency in the mental health field? Is it even “thinking” at all? To be sure, when employed—inadvertently or intentionally—it can effectively obscure research and treatment inadequacies, bully innovation, create cognitive shortcuts, and ineffectively deal with intolerably complex subject matter. Indeed, this tactic has caused consulting clinicians, sponsors, and the traumatized public to find it nearly impossible to make sense of this information, or to trust in those responsible for mental health care.

Binary thinking is the antithesis of dimensional thinking, and it is especially dangerous in medicine. It is the thinking of division. It falls short of even one dimension of complexity. For example, who can deny the ever-present tendency to identify politicians as either socialist or capitalist, an orbit circular or elliptical, and, as a cautious medical corollary, acute myocardial infarction with or without chest pain.

As an illustration, suicidal ideation has conventionally been considered the hallmark of significant and recognizable psychopathology. Hence, its presence identified by point A; absence point B, and written AB. This, of course, ignores dimensionality, which if accounted for, would look like: A . . . . . . . . . . B. In the correct formulation, not only are there intermediate points between A and B, but there are also points above and below the line.

. . . . . . . . . . . .

A . . . . . . . . . . B

. . . . . . . . . . . .

Therefore, in the vast space of suicidal ideation, there are many coordinates. It is possible to stratify ideational cohorts as fleeting, transient, impermanent, or enduring. Furthermore, ideation can vary according to age, ethnicity, gender, religious practice, cultural observance, psychopathology, childhood and family adversity, and country of origin. And, topical research indicates that the absence of ideation, like the absence of chest pain in women, is not a benign clinical finding. Consequently, the value of the independent finding of ideation is hard to determine.

Independent, measured, and patient-centered research data are critical to design future innovative research, provide evidence in the regulatory process, attract stakeholders, and modify and improve clinical practice. However, there continue to be significant problems in suicide investigation. Suicide studies have, for example, tended to focus on traditional, deep-seated, overrepresented constructs such as ideation, depression, hopelessness, and helplessness. Nonideation and unconventional states have been largely absent from investigation. Further, suicide studies tend to summarize large amounts of varied population-level information, which future researchers then rely on over time. The result is that stale ideation-centric hypotheses are re-applied in future studies. (Large et al., 2016)

The essential questions that need to be scrupulously and directly asked when evaluating any suicide investigation are:

1. What are the strict operational definitions of suicide, beyond intentional self-annihilation?

2. What variables define high and extreme risk patients, other than those identified in meta-analysis designs?

3. What factors, beyond depression and ideation, are involved in the study?

4. What are the impact of ideation and non-ideation states on attempts among suicide subgroups?

Perhaps most concerning is that suicide researchers, who ought to clearly understand the ethical dilemmas involved, have depended on meta-analyses as a shortcut to publication. Rather than ensuring thorough examination of critically ill persons—rather than “seeing,” examining, and differentially diagnosing—researchers have prioritized publication over things like: criteria for inclusion, selection of controls, detailed analysis, and careful reporting of results. In the words of William Osler, researchers have forgotten that “every patient you see is a lesson in much more than the malady from which he suffers.”

Of course, differences in studies will continue. But when the dissimilarities are large, confusing, and of low quality; when definitional domains are inadequate; when patient observations across disciplines are unconnected; and when out-of-date research methods are used, incoherency will reign (and adverse outcomes all but guaranteed).

Emergency psychiatric reality cannot be reduced to an either/or solution. Be suspicious of such simplistic suicide analyses, whatever the so-called distinguished sources. 


Reference: Copelan R. et al. Adolescent violence screening in the ED. American Journal of Emergency Medicine (2006)24;582-594.


Image: Lightspring / shutterstock


Russell Copelan, MD is a retired emergency department psychiatrist. He is a reviewer for the journal Academic Psychiatry.

More from Op-Med