The inevitable wastefulness of suicide prevention programs –
- The innocent and the cynical make emotional appeals for suicide prevention programs.
- Open-minded citizen-scientists ask for simple numbers to do some calculations. Obtaining these numbers, they come to inevitable conclusions.
- This blog post shows you how to analyze emotional appeals for suicide prevention programs, whether they are innocent or cynical.
A recent article in The Lancet Psychiatry
Khalifeh H, Hunt IM, Appleby L, Howard LM. Suicide in perinatal and non-perinatal women in contact with psychiatric services: 15 year findings from a UK national inquiry. The Lancet Psychiatry. 2016 Jan 16.
Was reviewed in the Mental Elf.
The blog post ended with a non sequitur:
Routine assertive follow-up should be provided to all women in the perinatal period to address the risk of suicide [emphasis in original].
The blog post received a number of tweets, including:
“This really needs to change urgently. Suicide during the perinatal period.”
“Suicide during the perinatal period… This is exactly why we need better services”
I got an expected response:
“… Are you suggesting some suicides okay?”
If you bring up evidence when you’re dealing with an emotional, politically charged topic, you at least get misunderstood. Maybe you get maligned and have to defend yourself.
A look at the numbers
Integrated data on all suspected suicides by women aged 16 – 50 between 1997 and 2012 in the UK with data concerning whether or not the woman had been in contact with psychiatric services in the 12 months preceding their death.
Reducing the sample to 4785 women in the perinatal period yielded 80 women who died by suicide during the first year after birth of the child and 18 who by suicide died during pregnancy.
So, we are dealing with approximately five women dying by suicide in the entire UK in the first year postpartum and one in the entire UK dying by suicide during pregnancy.
Statistical analyses with such small numbers should not be taken too seriously, but the blog post and the target paper refers to a reduced likelihood that women in the perinatal period who died by suicide were – compared to other women dying by suicide at other times –less likely to ever have been admitted to psychiatric facilities, less likely to be prescribed any psychiatric medication at the time of death, and less likely to be receiving psychotherapy. They were more likely to have a diagnosis of depression and a duration of illness less than a year.
With such small numbers of events – deaths by suicide – it makes much more sense to look at the circumstances of the few suicides than to try to speculate from multivariate statistics. But let’s go with what the article and blog present.
Women who died by suicide during pregnancy in the year afterwards are less likely to be admitted to inpatient units and less likely to be getting treatment at their end of life. We don’t know if the women made medical emergency room contacts as a result of attempted suicide, which is quite relevant. But it seems that some of the women were just coming into contact with the mental health system. Were they treated in primary care beforehand and only then referred onward when they became suicidal? We don’t know, but we need to know.
But more importantly, we’re talking a single death per year in all of the UK. Any loss of life is tragic, but are we prepared to reorganize services in the off chance that we could prevent this single death? In a time of scarce resources, what would be the trade-off? My understanding is that in the UK, referrals for psychotherapy can take longer to be completed than a pregnancy, particularly if the depression is detected in the third trimester.
Multivariate analyses with such small numbers are even more dubious. Nonetheless, when controls were introduced for age, ethnicity, marital status, and employment status, only a diagnosis of depression and an illness duration of less than a year remain significant.
Again, we’re really trying to make sense of unreliable numbers. But could it be that the few deaths by suicide among pregnant women and women who recently gave birth had unwanted pregnancies? Were there issues associated with partner or family reactions to the unwanted pregnancy contributing to wish to die?
Would the system be better off focusing on how unwanted pregnancies are dealt with than with creating specialty programs to prevent suicide among these women?
Effectively dealing with an unwanted pregnancy could be a benefit in itself, aside from whether it reduced the single death by suicide among pregnant women per year in the UK or the five in the first year postpartum.
Having been a scientific advisor to over a decade of futile suicide prevention programs, I’m convinced that it is better to create programs which are justified by sufficient evidence for meeting goals other than reducing suicide. We might hope that they reduce the infrequent event of death by suicide, but don’t reasonably expect to demonstrate that.
More complex multivariate analyses take us far into the realm of voodoo statistics. Nonetheless, when additional controls were introduced such as depression diagnosis, alcohol misuse, drug issues, personality disorder, and history of psychiatric admissions, no risk factors remain significant.
The naïve and innocent
Another Mental Elf post discussed the article:
Bolton JM, Gunnell D, Turecki G. Suicide risk assessment and intervention in people with mental illness. The BMJ 2015;351:h4978
The abstract of the article is:
Suicide is the 15th most common cause of death worldwide. Although relatively uncommon in the general population, suicide rates are much higher in people with mental health problems. Clinicians often have to assess and manage suicide risk. Risk assessment is challenging for several reasons, not least because conventional approaches to risk assessment rely on patient self reporting and suicidal patients may wish to conceal their plans. Accurate methods of predicting suicide therefore remain elusive and are actively being studied. Novel approaches to risk assessment have shown promise, including empirically derived tools and implicit association tests. Service provision for suicidal patients is often substandard, particularly at times of highest need, such as after discharge from hospital or the emergency department. Although several drug based and psychotherapy based treatments exist, the best approaches to reducing the risk of suicide are still unclear. Some of the most compelling evidence supports long established treatments such as lithium and cognitive behavioral therapy. Emerging options include ketamine and internet based psychotherapies. This review summarizes the current science in suicide risk assessment and provides an overview of the interventions shown to reduce the risk of suicide, with a focus on the clinical management of people with mental disorders.
I agree that the most compelling evidence is for lithium in the case of bipolar disorder and cognitive behavior therapy for depression, but the evidence is also sufficient to be providing those treatments for bipolar disorder and depression, respectively. Any reduction in suicide is a plus.
The blog post tries to spin a hopeful picture:
Emerging methods of supporting a clinical assessment (e.g. the Implicit Association Test), which have an empirical basis, and are interview independent, may provide an important new avenue for supplementing clinical suicide risk assessment. Whilst the review may paint a somewhat bleak picture of the ‘state of the art’ of suicide risk assessment, this is not a message of hopelessness, but rather a clarion call to action for further research into how we can refine risk assessment and intervention development.
I really don’t think the Implicit Association Test is going to get around the many persons not expressing intention to die by suicide while in contact with mental health or other professionals. I seriously doubt whether it has been validated as a predictor of suicide, rather than a surrogate end point like suicidal ideation.
The blog post ends with a realistic appraisal of the difficulty predicting suicide but slides into another emotional non sequitur:
Given the fluctuating nature of suicide risk and the fact that it is a rare event which cannot accurately be predicted, the only safe response is to take all suicidal thoughts seriously and respond appropriately. NICE advocate a needs and assets-based approach after self-harm rather than focusing on risk. We welcome the day when everyone at risk of suicide is responded to with compassion, confidence, and competence, and have a co-created safety plan to ensure their safety.
The Mental Elf blog post ended with an emotional plea. But what does this mean and what action does it call for? How practical is it? These questions need to be asked.
The cynical and opportunistic
The Mental Elf blog post claims that women who discontinue antidepressant medication during pregnancy have a significant risk of depression versus those who continue, citing this article:
Cohen LS, Altshuler LL, Harlow BL, Nonacs R, Newport DJ, Viguera AC, Suri R, Burt VK, Hendrick V, Reminick AM, Loughead A. Relapse of major depression during pregnancy in women who maintain or discontinue antidepressant treatment. JAMA. 2006 Feb 1;295(5):499-507.
The Cohen article created a scandal and an embarrassment for JAMA. The lead author had undisclosed conflicts of interest – ties to the pharmaceutical industry – and oddly biased and unrepresentative sampling yielded an exaggerated risk within this study. The sample was certainly not representative of the typical woman on antidepressants when she discovers she is pregnant. The risk was exaggerated.
JAMA was so embarrassed by being snookered by this and other articles by psychiatrists with undeclared conflicts of interest that in the aftermath, an editor declined our proposal to write a systematic review of screening for depression during the perinatal period. This was confirmed in a number of phone calls with the editor. Our article had to go elsewhere, landing in a more obscure place, after evaluation by reviewers who seem to have no familiarity with depression in the perinatal period:
Thombs BD, Arthurs E, Coronado-Montoya S, Roseman M, Delisle VC, Leavens A, Levis B, Azoulay L, Smith C, Ciofani L, Coyne JC. Depression screening and patient outcomes in pregnancy or postpartum: a systematic review. Journal of Psychosomatic Research. 2014 Jun 30;76(6):433-46.
The issue of whether a pregnant woman should initiate or continue antidepressants is a complex, personal decision concerning risk versus benefits that takes other risks into account, including past history of depression and suicide attempts. Such a discussion deserves an informed discussion with a primary care physician, who seldom are willing or prepared to offer them the time.
The naïve and the cynical: David Cameron pledges money for a revolution in maternal mental health
David Cameron won’t be remembered for improving mental health services. . But he scores a lot of points by proposing the creation of specialized services – without really addressing the serious crisis in routine provision of mental health services to all – inaccessibility and long delays to actual receipt of services.
The prime minister will pledge on Monday to end the postcode lottery under which three-quarters of the 40,000 women a year who experience conditions such as postnatal depression do not receive vital treatment intended to keep families together, protect babies and reduce the risk of maternal suicide.
I think what we have discussed above should raise skepticism about whether the specialized services will actually reduce maternal death by suicide.
Pledging a huge expansion of services to tackle the huge unmet need in maternal mental health, Simon Stevens, chief executive of NHS England, told the Guardian: “At the moment about 40,000 women who are pregnant or within the first year of having their baby have a severe mental health problem. But of those 40,000 only about 10,000 are at the moment getting access to specialist perinatal mental health services. Three out of four are missing out. But by the end of the decade we are going to make that a universal offer, so all 40,000 will get access to a local specialist team.”
A tug at our heartstrings but let’s watch our wallets.
If we’re talking about “severe mental health problems,” then we’re mostly talking about pre-existing conditions, most of which have been treated within routine care. The big challenge is not letting receipt of those services get disrupted by pregnancy and care for an infant or restoring the connection if indeed these services are disrupted. Given the prevalence of “severe mental health problems” and the difficulties getting pregnant women and mothers of infants to their appointments, it would seem best to facilitate strengthening their connection with pre-existing care.
Guaranteeing care for every mother who needs it by 2020 will tackle what the Maternal Mental Health Alliance – a coalition of more than 60 organisations that work on the issue – claims are “shocking gaps” in “patchy” NHS maternal mental health services.
In his speech, Cameron will announce that the NHS will put £290m into creating new community perinatal mental health teams and more beds in mother-and-baby units. They help women battling post-traumatic stress disorder, postpartum psychosis and other serious similar conditions. There are about 120 in England but experts such as Andy Bell, deputy director of the Centre for Mental Health, say 60 more are needed.
Postpartum psychosis is dramatic and carries risk to the infant as well as the mother. But it is fortunately rare, and has different causal factors than more common major depression. A good predictor is a past psychotic experience.
Suicide is the second biggest cause of maternal death after sepsis, a violent immune reaction caused by serious infection. A recent major inquiry into such deaths found that mental health problems were involved in 23% of them and that one in seven is from suicide. More than 100 suicides occurred between 2009 and 2013, it found.
Here we go again: we need to consider absolute rates of death by suicide, not relative rates, deaths are relatively uncommon during pregnancy in the first year postpartum.
Dr Dan Poulter, the Conservative MP who was minister for maternity care until last May, said: “It’s frankly unacceptable that mental health is the second commonest cause of maternal death.”
Dr. Poulter, exactly how is it unacceptable and do you really think that anything you do will make a difference? What exactly are you doing by declaring it “frankly unacceptable” except for looking good politically?
“Mental health has been the Cinderella of the NHS and perinatal mental health services haven’t historically received the investment attention they needed. If they’d had that, we’d have been able to identify more mums at risk and prevent many of these deaths from suicide.
Dr. Poulter, you are naïve or cynical or both.
In his highly political role as President of the Royal College, Simon Wessely is hardly in a position to counter emotion with evidence. Instead, he is evasive and vaguely goes along with the emotion:
Prof Sir Simon Wessely, President of the Royal College of Psychiatrists, said: “Suicide in anyone is a tragedy, but the impact on a new family is probably as bad as it gets, so extending quality mental health to all, and not just some, of those new mothers with serious depression or psychosis is clearly the right thing to do.”
Yep, probably as bad “as it gets” and so develop any special services is “clearly the right thing to do.”
As I was uploading this blog post, I came across highly relevant criticism of the David Cameron’s government being out of touch with the evidence:
David Cameron’s government is failing to collect the vital data needed to help tackle Britain’s growing mental health care crisis, the Mirror reveals.
Bungling Tories admit they have no idea how many young parents have committed suicide in the UK – nor how many people diagnosed with mental illness end up in prison.
Labour will highlight 30 different mental health related issues on which the Tories were unable to provide any information in a hard-hitting #mentalhealthmatters campaign.
Shadow Minister for Mental Health Luciana Berger last night branded the situation “absolutely appalling.”
Andy Bell, deputy chief executive at the Centre For Mental Health said: “It is surprising what we don’t know.”
Suggestions for dealing with the inevitable future emotional appeals for suicide prevention.
- Ask about absolute numbers of how many lives would be saved if everything went as planned.
- Ask about the likelihood of these lives actually being saved given the unpredictability and infrequency of death by suicide, no matter how tragic.
- Ask about the specialized services that would remain unutilized or underutilized if they are devoted to dealing specifically with risk of suicide.
- Ask about what mental health needs of persons at high risk of suicide would be better addressed within existing mental health services, emphasizing continuity of care, rather than setting up separate specialized services.
- Ask what improvements to the existing inadequacy of mental health care would be postponed if the services were diverted on the basis of emotion, not evidence.
Specifically for suicide during the peripartum –
- Ask whether those who demand specialized programs seem aware that depression is now more common among pregnant and peripartum women versus other woman of the same age and suicide is less common.
- Ask whether addressing resolution of unwanted pregnancies, including those due to rape or incest, would be a more efficient use of resources in terms of overall improvement in maternal well-being than committing the same resources specialized services attempting to prevent infrequent suicides.
Why is the topic of unwanted pregnancies so seldom raised in discussions of perinatal suicide?