Preventing suicide in all the wrong ways

My introduction to a guest post by D.J. Jaffe

JaffeDJ Jaffe is Executive Director, Mental Illness Policy Organization and his op-eds appear in the New York Times, Washington Post, Sacramento Bee, New York Daily News, New York Post, San Diego Union Tribune, Albany Times Union and numerous other publications. Read more about him here. You can follow him on Twitter, @MentalIllPolicy.

I really like what D.J wrote because I am a battle-weary veteran of over a decade as an external scientific adviser on ambitious, multilevel interventions to reduce suicidality and depression on a regional basis in Europe.

The sad reality is that population-based suicide interventions program just can’t be expected to work if we measure their efficacy in terms of reduction in suicide. Because suicide is a rare event, a population of 500,000 would not be sufficient to detect a statistically significant reduction in suicide rates of less than 30%.

If a study focuses on a population of a half million people, a lot goes on at the same time as the intervention over which you have no control. It will be very difficult to determine if differences between an intervention and a control region are actually due to the intervention.

The last European project in which I was involved had intervention and control regions in each of four countries. In one of the countries, there were large scale floods in the intervention but not the control group. In another country, there was a serious decline in the economy affecting both the intervention and control regions, the effects of which would expect to dwarf any effects of intervention.

Before that series of projects, I was a reluctant coinvestigator on PROSPECT, a NIMH project to reduce suicide. The project was doomed to uselessness from the start, but its funding was dictated by a US Senator grieving the loss of his father who died from suicide after visiting a primary care physician.

The guiding rationale for PROSPECT study was that elderly males are a particularly high risk group for suicide, but they have low uptake of depression treatments. Often, like the senator’s father, elderly males who died by suicide visit a primary care physician in the months before their death, but without the risk ever being detected. So, we need to evaluate an intervention to improve the detection and treatment of depression in the elderly in primary care. Our estimates at the time was that although the elderly constitute 13% of the population, they represent 18% of deaths by suicide.

Even among this (modestly) high risk group, suicide is such an infrequent event, we cannot reasonably expect to register a reduction in deaths by suicides as a result of intervention. So, PROSPECT targeted suicidal ideation as a surrogate endpoint. The problem with this particular target is that most persons with suicidal ideation will not attempt to harm themselves. If you obtain measures of suicidal ideation and track a large sample over time until some deaths by suicide accumulate, most of those deaths will not be among the highest scores on the measure of suicidal ideation.

Also, measures of suicidal ideation are more related to measures of depression then to the likelihood of death by suicide, or even suicidal behavior. So, in the case of PROSPECT, we were left with a study essentially focused on improving detection and treatment of depression in primary care. We already have a lot of such studies in literature, and that’s not what the good senator wanted, so, PROSPECT was dressed up as a study of prevention of suicide.

But, even with a torturing of the data, results were not impressive. Here is the primary outcome paper. Here is my critique. I could have said more. For instance, the study involved 20 primary care practices across three regions of the Northeastern United States. Differences in suicidal ideation between regions, which included both intervention and control groups, were greater than any differences between intervention and control group in suicidal ideation. Another complex intervention gone bad.

What if a skeptic says ‘I’m not impressed by surrogate measures, I want to know if this multimillion dollar project actually reduced suicide. Where is the evidence?

Unfortunately, the skeptic would be out of luck with this reasonable question:

One patient in the intervention group died by suicide; the follow-up methodology did not permit us to know reliably the causes of death for patients in the usual care group. Two patients, 1 in the intervention group and 1 in the usual care group, made suicide attempts.

 The urge for yet another large suicide prevention effort often comes on the heels of a well-publicized event like suicide of a celebrity or an apparent cluster of teenage suicides in a community.

As infrequent events, rates of suicides often fluctuate for unknown reasons. A small increase can seem large when expressed as percentage. 300% is alarming until you know the baseline was 1 suicide.

 I was involved in a one-country study that appeared to produce a reduction in the composite outcome of deaths by suicide plus suicidal acts of high lethality – hanging, shooting, and jumps off high places. There appeared to be a reduction in these variables in the year after intervention. However, the reviewers of the manuscript reporting the study asked that the data be obtained for a year prior to the intervention and then a year after the follow-up. When that was done, the differences that were going to be attributed to the intervention were less than the usual fluctuation over time. The apparent effects of the intervention may have been a fluke.

 Outside of a controlled study, we can expect that most increases in deaths by suicide will be followed by decreases as result of regression to the mean. Often communities then congratulate themselves on what their goals having been accomplished.

 Too often, suicide interventions are driven by the idea that we just cannot not do something about such a pressing problem. I think we need to reflect on that issue before we undertake any more ambitious initiatives.

suicide preventionBelow is an excerpt from a large article by DJ Jaffe. I encourage you to access and read the whole article, Preventing suicide in all the wrong ways.

DJ Jaffe writes with a sometimes irritating honesty. But I just can’t find anything to which I object.


Here are the points to be noticed in the excerpt below:

  • There is little scientific evidence media campaigns reduce suicide and mounting evidence they don’t.
  • Mental health industry sponsored suicide initiatives are often targeted at college students, a group least likely to commit suicide. 
  • Why are mass market media campaigns so popular in spite of the fact there is no evidence they work and evidence they don’t? Money.
  • One effective suicide prevention strategy is means removal: putting locks on guns, medicine cabinets and drawers containing knives.
  • One effective suicide prevention strategy is means removal: putting locks on guns, medicine cabinets and drawers containing knives.
  • If the mental health industry insists on relying on communications as a path to reduce suicide, then those communications would be much more effective if they were targeted at those of highest risk of suicide, not the general public.

Suicide cannot be reduced through awareness advertising and public relations

Every suicide is a tragedy for the individual, their family, and the community. Many of the truly mental illness related suicides could be prevented if persons with mental illness were provided care. Instead of doing that, the mental health industry’s main tool in reducing suicide takes the form of public service announcements, brochures, hotlines, and speeches targeted to the general population. For example, in 2012, the California mental health industry banded together to spend $32 million in public funds for a TV, radio, billboard, online, mobile and print advertising campaign targeted at the general public to reduce suicide. (California Mental Health Services Authority 2012).  But those charged with overseeing the funds, refuse to measure rates of suicide to see if the funds are having an impact. Instead they measure tangential issues like “attitudes” and number of presentations made. The money is wasted.

There is little scientific evidence media campaigns reduce suicide and mounting evidence they don’t. The largest and most sound review of the issue was Suicide Prevention Strategies: A systematic review, published in the Journal of the American Medical Association.  (J. John, Alan and al. 2005). The authors found that

despite their popularity as a public health intervention, the effectiveness of public awareness and education campaigns in reducing suicidal behavior has seldom been systematically evaluated.

The report went on to note what the research does show:

Such public education and awareness campaigns, largely about depression, have no detectable effect on primary outcomes of decreasing suicidal acts or on intermediate measures, such as more treatment seeking or increased antidepressant use.

A 2009 study in the journal Psychiatric Services looked at 200 publications between 1987 and 2007 describing depression and suicide awareness programs targeted to the public and found that the programs “contributed to modest improvement in public knowledge of and attitudes toward depression or suicide,” but could not find that the campaigns actually helped increase care seeking or decrease suicidal behavior. A similar study in 2010 in the journal Crisis actually found that billboard ads had negative effects on adolescents, making them “less likely to endorse help-seeking strategies”. (Sanburn 2013)

Mental health industry sponsored suicide initiatives are often targeted at college students, a group least likely to commit suicide.  The 2011 National Survey on Drug Use and Health is one of the premiere epidemiological surveys and found college students were less likely than other same aged adults to have serious thoughts of suicide (6.5 vs. 8.4 percent), make suicide plans (1.5 vs. 2.4 percent), or attempt suicide (0.8 vs. 1.8 percent). (SAMHSA 2012). The college targeted PR programs are no more effective than mass market anti-suicide PR campaigns.

Few such programs are evidence-based, reflect the current state of knowledge in suicide prevention, or evaluate effectiveness and safety for preventing suicidal behavior…A systematic review of studies published from 1980-1995 found that knowledge about suicide improved but there were both beneficial and harmful effects in terms of help-seeking, attitudes, and peer support.” (J. John, Alan and al. 2005)

Why are mass market media campaigns so popular in spite of the fact there is no evidence they work and evidence they don’t? Money. It is very easy and profitable for a mental health provider to write a brochure, produce a PSA, rather than try to reduce suicide. By putting their logo on the materials they increase their visibility and self-importance. As one suicide researcher concluded, “The conflict between political convenience and scientific adequacy in suicide prevention is usually resolved in favor of the former. Thus, strategies targeting the general population instead of high-risk groups (psychiatric patients recently discharged from hospital, suicide attempters, etc.) may be chosen…especially if the desired outcomes also include a number of conditions frequently associated with suicidal behaviors (such as poor quality of life, social isolation, unemployment and substance misuse).” (Diego de Leo 2002)

How to reduce suicide

One effective suicide prevention strategy is means removal: putting locks on guns, medicine cabinets and drawers containing knives. (Yip, et al. 2012). However, the mental health industry has largely been unwilling to give up funds they can use to create TV ads featuring their logo in order to fund suicide means reduction. California did authorize the use of mental health dollars to fund a net under the Golden Gate Bridge. But that was largely a PR ploy to defuse criticism of massive waste in California’s Mental Health Services Act (MHSA) fund which is supposed to fund services for the seriously ill. (Mental Illness Policy Org. August, 2013) California Senate President Pro Tem Darrell Steinberg claimed “Proposition 63’s contribution to suicide prevention at the Golden Gate Bridge will probably become its most publicly recognizable benefit.” (Steinberg 2014).

It is also known and ignored that those who are most likely to commit suicide are those who have previously attempted suicide, first-degree relatives of those who completed suicide, and persons with serious mental illness. (Tsuang 1983), These individuals, by name, are likely known to the mental health system as a result of their suicide or family histories. Intensive follow up of these individuals, rather than the general public, would be a much more efficient and effective way for the industry to reduce suicide. Time Magazine reported on this in an interview with Lanny Berman, executive director of the American Association of Suicidology (AAS):

The general zeitgeist in the field is public education is good, and it’s better that people know about the problem and really know that prevention is possible. But I don’t know that public awareness campaigns work for the people you most want to reach, the people who are already suicidal.”  If we know who’s most at risk, people like Jaffe and Berman argue, shouldn’t we target them in a smarter way? If a factory closes, for example, shouldn’t efforts be made to market suicide prevention services in that community? …Berman… is concerned that SAMHSA is too focused on “upstream” measures like increasing overall awareness. “The bottom line is that the people most at risk are people who don’t get into treatment, and a public health approach shifts attention from high-risk patients to large populations of folks who might develop mental health problems,” he says. (Sanburn 2013)

While the lifetime risk for suicide in people with schizophrenia is only 5%, we do know how to predict and prevent those suicides. The biggest risk factors are “number of prior suicide attempts, depressive symptoms, active hallucinations and delusions, and the presence of insight…a family history of suicide, and comorbid substance misuse. The only consistent protective factor for suicide was delivery of and adherence to effective treatment.” (Hor and Taylor 2010)

As far as we know, no mental health provider is proposing to use suicide funds to treat the seriously ill. There is evidence they should.  A Kendra’s Law study in New York found Assisted Outpatient Treatment, reduced suicide attempts and physical harm to self 55%. (New York State Office of Mental Health 2005). The Treatment Advocacy Center compiled a list of suicide studies suggesting suicide is more likely to occur in those individuals with schizophrenia and bipolar disorder who are not being treated or adequately treated and less likely in those that are treated.

  • A 34-year follow-up study of 158 individuals with bipolar disorder reported that 18 of them (11 percent) had committed suicide. The suicide rate was more than twice as high among patients who had not been treated compared with those who had been treated (p = 0.04), a difference the authors called “spectacular.”
 (Angst F 2002)
  • A study in Kentucky found that only 2 of 28 individuals with schizophrenia who committed suicide had evidence in their blood of having taken antipsychotic medication. Thus, 93 percent of them were not being treated. (Shields LBE 2007)
  • A case control study of 63 individuals with schizophrenia who committed suicide and 63 individuals with schizophrenia who did not reported that “there were seven times as many patients who did not comply with treatment in the suicide group as there were in the control group.”
 (De Hert M 2001)
  • A case-controlled study compared 27 inpatients with schizophrenia and 24 inpatients with affective psychoses, all of whom suicided, with their matched inpatient case controls who did not suicide.  The authors concluded that there is “a significantly increased risk” of suicide when medications are not used. (Gaertner I 2002)

If the mental health industry insists on relying on communications as a path to reduce suicide, then those communications would be much more effective if they were targeted at those of highest risk of suicide, not the general public. Prisoners are a high-risk group. Suicide in jail is three times more common than in the general population and thirty eight percent of those who committed suicide in jail in 2005-2006 had a known history of mental illness. (Hayes April, 2010). If the mental health industry worked to reduce incarceration of persons with mental illness, they could further reduce suicide.


To reduce suicide we should stop funding what doesn’t work and start funding what does. The Helping Families in Mental Health Crisis Act (HR-3717) proposed by Representative Tim Murphy (R-PA) along with 96 bipartisan co-sponsors takes a step in this direction by ensuring mental illness spending is driven by evidence and through funding various anti-suicide initiatives.

Works Cited

American Foundation for Suicide Prevention. Suicide: Facts and Figures. 2010. (accessed April 15, 2013).

Angst F, Stassen HH, Clayton PJ et al. “Mortality of patients with mood disorders: follow-up over 34-38 years.” Journal of Affective Disorders 68 (2002): 167-181.

California Mental Health Services Authority. California Mental Health Services Authority Launches Statewide Suicide Prevention Campaign. Dec 12, 2012. (accessed 12 28, 2013).

Center for Disease Control and Prevention. “Surveillance for Violent Deaths — National Violent Death Reporting System, 16 States, 2010 .” Morbidity and Mortality Weekly Report, January 17, 2014: 1-33.

De Hert M, McKenzie K, Peuskens J. “Risk factors for suicide in young people suffering from schizophrenia: a long-term follow-up study.” Schizophrenia Research 47 (2001): 127-134.

Diego de Leo, Franzcp. “Why are we not getting any closer to preventing suicide?” The British Journal of Psychiatry 181 (2002): 372-374.

Gaertner I, Gilot C, Heidrich P. “A case control study on psychopharmacotherapy before suicide committed by 61 psychiatric inpatients .” Pharmacopsychiatr 35 (2002).

Hayes, Lindsay M. National Study of Jail Suicide: 20 Years Later. National Institute of Corrections, U.S. Department of Justice, Available at, April, 2010.

Hor, Kahyee, and Mark Taylor. “Suicide and schizophrenia: a systematic review of rates and risk factors.” Journal of Psychopharmacology, November 2010: 81-90.

John, Mann MD, Apter MD Alan, and et. al. “Suicide Prevention Strategies  Systematic Review.” The Journal of the American Medical Association (JAMA) (American Medical Association) 294, no. 16 (October 2005): 2064-2074.

Mental Health America. Suicide. (accessed April 15, 2013).

Mental Illness Policy Org. “California’s Mental Health Service Act: A Ten Year $10 Billion Bait and Switch.” New York, NY, August, 2013.

National Alliance on Mental Illness (NAMI). Suicide Fact Sheet. January 2013. (accessed April 15, 2013).

New York State Office of Mental Health. Kendra’s Law: Final Report on the Status of Assisted Outpatient Treatment. Report to Legislature, Albany: New York State, 2005, 60.

SAMHSA. “Results from the 2011 National Survey on Drug Use and Health: Mental Health Findings.” Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration , Rockville, MD, 2012.

Sanburn, Josh. “Inside the National Suicide Hotline: Preventing the Next Tragedy.” Time, September 13, 2013.

Shields LBE, Hunsaker DM, Hunsaker JC III. “Schizophrenia and suicide: a 10-year review of Kentucky Medical Examiner cases.” Journal of Forensic Sciences 52 (2007): 930-937.

Steinberg, Darrell. Steinberg on vote for Suicide Barrier at Golden Gate Bridge. June 27, 2014. (accessed July 20, 2014).

Torrey, E.F. 5000 suicides a year are likely caused by mental illness, mainly untreated mental illness. . (accessed 4 15, 2013).

Tsuang, MT. “Risk of suicide in the relatives of schizophrenics, manics, depressives, and controls. T.” Journal of Clinical Psychiatry 44, no. 11 (November 1983): 398-400.

Yip, Paul, Eric Caine, Saman Yousuf, and Shu-Sen Chang. “Means restriction for suicide prevention.” Lancet 379, no. 9834 (June 2012): 2393 – 2399.