Saturday, March 27, 2004

Antidepressant Medication and Suicide
A Meta-Review of the Blogosphere

There is concern that there could be a link between antidepressant drugs and suicidal thoughts and behavior.  This is something that has generated a lot of commentary in the blogosphere:
Waypath buzz-o-meter-graph
I would like to take some time to pull together the recent medical information, news reports, and blogger commentary; then provide my own views on the subject.  I conclude with some recommendations.  If you already are familiar with this topic, just skip down to the recommendations. 

Starting with the medical information, the Canadian Medical Association Journal has three articles of interest:

  1. UK psychiatrists question SSRI warnings for under-18s

  2. Antidepressants and adverse effects in young patients: uncovering the evidence

  3. Facing the evidence: antidepressant treatment in children and adolescents

These articles are in the Commentary section.  As such, they are not original, peer-reviewed research articles.  I included them because the perspective of persons in Canada is likely to be different from that seen in the USA.  They have a single-payer health system rum by the government.  The effect of politics, finance, and lobbying will be different in Canada.  See the bibliography in the CMAJ articles for links to the original research that has prompted all this buzz.

The FDA has issued reportss:

  1. Search results from www.fda.gov

  2. FDA Public Health Advisory dated October 27, 2003

  3. FDA Public Health Advisory dated March 22, 2004

  4. FDA Talk Paper dated  March 22, 2004

  5. Background Information on the Suicidality Classification Project

Not all of the recent scientific articles about antidepressants and suicide have concluded that there is a risk; some have indicated a reduction of risk::

  1. association between antidepressant prescribing and rates of suicide from 1991 to 2000Relationship Between Antidepressant Medication Treatment and Suicide in Adolescents (Arch Gen Psychiatry. 2003;60:978-982.)

  2. Association between antidepressant prescribing and suicide in Australia, 1991-2000: trend analysis (BMJ 2003;326:1008-1011)

  3. also see the BMJ editorial Unknown unknowns in suicide and depression

  4. Suicide Rates in Clinical Trials of SSRIs, Other Antidepressants, and Placebo: Analysis of FDA Reports (Am J Psychiatry 160:790-792, April 2003)

  5. also see the POEM analysis in AFP: Suicide Rates in Patients Taking Antidepressants 

  6. Prospective Study of Fluoxetine Treatment and Suicidal Behavior in Affectively Ill Subjects (Am J Psychiatry 156:195-201, February 1999)

  7. Symptom Reduction and Suicide Risk in Patients Treated With Placebo in Antidepressant Clinical Trials (Arch Gen Psychiatry. 2000;57:311-317)

  8. Reduction by Paroxetine of Suicidal Behavior in Patients With Repeated Suicide Attempts But Not Major Depression (Am J Psychiatry 155:543-547, April 1998)

There are many news articles and editorials; representative examples include:

  1. Fox News: (AP) FDA Wants Suicide Warning Labels on Antidepressants, Monday, March 22, 2004WaPo News: New Warning Urged On Antidepressants, Tuesday, March 23, 2004

  2. WaPo Editorial: Keeping Doctors in the Dark,  March 24, 2004

  3. Ambulance-chaser press release on eMediaWire

  4. (Psychology Today) Antidepressants: The Kid Question, Mar 8, 2004 

From the blogosphere:

  1. From Depression to Suicide?

  2. Suicide Risk? Hrmmm....

  3. Health: FDA decision harms consumers

  4. Tough Times At GSK

  5. Antidepressants and Suicide: Where's the Culpability of the Doctors As Well?

  6. Less Than Zero

  7. Damned if you do, damned if you don't

  8. A Bitter Irony

  9. Oxymoron of the Day: Antidepressants May Cause Depression

  10. Federal Department of Asininity

  11. The danger of anti-depressants

  12. Warning: May Cause Suicide


  14. Politics at Your MD

  15. FDA issues suicide caution for antidepressants

There are others, of course, but I could spend all day tracking them all down.  I stopped when I got the impression that I was unlikely to find many more new ideas.

None of what I have to say will make any sense unless you have wither been following this topic closely, or you have read at least one good summary paper.  If neither is true for you, I suggest you scan the article #2 in the first group: Antidepressants and adverse effects in young patients: uncovering the evidence.

Summary:  drawing from the links above, and my persona experience, I would like to present a summary of all of this.   It appears that the opinion in the blogosphere is split.  Some persons feel that the suicide risk is an example of a cover-up, that it is a reason to not prescribe antidepressant, or that it is a reason to turn to alternative medicine instead of refined pharmaceutical.  Many people express concern about the influence of financial concerns when it comes to marketing and regulating medications.  Some express the belief that the FDA is being overly zealous in requiring warnings about suicide risk.  Some people point out the difficulties of making sense of all the available information on the link between antidepressants and suicide.  Most notable are the articles from The Toxicology Weblog (#1 above) and Blogcritics (#3).  One commentator speculated that the FDA issued the warnings in order to help shield pharmaceutical companies from litigation (#14). 

In an attempt to get a clear picture of how best to think about this topic, I thinks it is good to get a sense of the fact that what we are talking about is the probability that something bad could happen as a result of trying to make something good happen.  It is important to recognize that, for persons with major depressive disorder, there is no risk-free option.  That is, there is going to be a risk of suicide no matter what you do.  All you can do is alter the probabilities. 

For persons with MDD, there is a risk that they could commit suicide on and off medication, and that if on medication, that the suicide could be caused by the medication or by something else.  Note that this is an oversimplification.  In most cases, suicide is a complex event.  There are multiple factors that lead to suicide in an individual, and in practice it is going to be very difficult to identify a singular cause.   Despite that complication, it still is useful to enumerate the possible risk scenarios:

Person with MDD, off medication:

  • Risk of suicide due to disease

  • Risk of suicide unrelated to disease

Persons with MDD on medication:

  • Risk of suicide due to disease

  • Risk of suicide due to medication

  • Risk of suicide unrelated to disease or medication

Note that, in the unmedicated group, there will be a subpopulation of people who do kill themselves, who would not have done so if they had been on medication.   Keep in mind that the goal here is to minimize total risk,  because it is the total risk that affects the patient's prognosis.  Thus, if you do something to increase one risk, but also lower the risk in a different area, the total risk may go down.  Thus, even if antidepressant use can increase suicide risk in some situations, it might lower the risk in other situations, with a new effect being that overall risk is lowered.  Note also that that the risk of suicide due to antidepressant medication, in the unmedicated group, is zero (by definition).  Therefore, if there are any  suicides related to medication in the medicated group, the risk of suicide due to medication will not be zero. 

Some commentators have wondered how it could be that we do not actually know the relative risk for these subpopulations.  But think for a moment about what you would have to do to get really solid information.  First of all, you would have to do a very large, randomized, placebo controlled study.  To average out the effects of age, health care delivery system, geography, and other confounding variables, you would have to have a multi-site study, preferably including centers in several different countries.  You would have to keep close track of thousands of people all over the world for at least several months; years would be better.  You also would have to have some objective way to tell which suicides in the medicated group were due to the medication.  It is not obvious that there is any way to do that.

Such a study would be enormously expensive.  Pharmaceutical companies could sponsor it, but no one would accept the result uncritically in that case.  It would have to be sponsored by the World Health Organization, NIMH, on some other neutral agency with deep pockets.  Would this be the best use of the resources needed to conduct the study?  Remember, if this study is done, that would mean resources would not be going toward understanding HIV, SARS, cancer, etc.  Suicide is a fairly common cause of death, but it is not in the top ten causes.  Therefore, it is not likely that the definitive study will be done anytime soon.

Therefore, we are left having to make important -- life changing -- treatment decisions, while we know full well that we do not have the information we need to accurately judge the relative risks of the treatment vs. no treatment.

One thing I noticed, especially in reading the blogs, magazine articles, and editorials, is that a lot of people have really strong feelings about the topic.  In some cases, it appears that people equate the issue of suicide risk due to antidepressants with the global question of whether antidepressants are good or bad.  This is somewhat like using the alleged breast cancer - abortion link as a proxy battleground for the pro-choice/pro-life debate.  I submit that this fuzzy overgeneralization is not helpful.   It only confuses the issue. 

There is another source of potential fuzzy thinking that should be clarified.  Let us say that we do somehow find a way to quantify the pertinent relative risks.  What would we do with the information.  Remember, the relative risks are valid for large populations.  They may or may not be applicable to any given individual. 

Also, remember that the statistics could be used as a guide to develop public policy that is good for the population as a whole, but not necessarily to any given individual.  A good example of this is pertussis vaccine.  If everyone gets the vaccine, the population is better off.  But some individuals will have serious reactions.  Because there can be a difference between what is good for the group and what is good for the individual, we could end up seeing people promoting the wider use of antidepressants on public health grounds, while a separate group argues against it on for individual reasons.  This is a kind of an argument that is difficult to resolve. 

One thing that is apparent from the articles cited above is the there may bee an important difference in the relative risk for adults compared to children and adolescents.  Most commentators agree that we need to be careful especially when it comes to treatment of children.   Unfortunately, there is a severe shortage  of trained child psychiatrists.  In the USA, there is a dearth of hospital beds for young persons with mental illness.  Even when hospitalization is an option, it is difficult to keep the patient in the hospital for more than a few days.   Also, most children are limited in their capacity to understand what is happening to them, and to report problems when they should.  For these reasons, we need to be especially careful about exposing children to medication that could have serious risks. 

Some people have been critical of the FDA for its stance that in cannot render a definitive statement about the risk of antidepressant-related suicide risk.  However, seeing that there is a tremendous amount of data, some of it contradictory, I think the FDA is correct to refrain from making a final pronouncement. Another criticism of the FDA has been that they do not have a lot of specific recommendations.  WHile they do post several recommendations, they are all general tips that represent basic good medical practice.  It is not clear that they are recommending anything other than a vague 'be careful' warning. 

My recommendations:  I would like to see a big multi-center study set up to try to resolve the question of possible antidepressant-induced suicide risk.  If the study is done, regardless of whether it shows an increased or decreased risk, or no effect, it is likely to conclude with a recommendation for more careful attention to premarketing drug trials; publication of articles with negative results, just as prominently as those with positive results; and closer follow-up of persons treated with antidepressant medication.  All of this will cost money, of course.  There are some simple, practical things to do in the meantime:

  1. If a patient calls to make an appointment for evaluation of depression, they can try to schedule two or three appointments all at once, at weekly intervals.  The first would be for an evaluation, the subsequent ones would be for follow-up.  In actual practice, it often is hard to schedule a follow-up appointment for just one week after the initial assessment.  If you schedule both up front, you are sure to be able to get in within a week of starting medication. 

  2. Any patient who gets an Rx for an antidepressant should be instructed in how to make contact in emergencies.   They also should prepare themselves mentally for the fact that, if there is a crisis, they may need to go to an emergency room or talk to an on-call doctor whom they do not know.  No one likes to do this, but if one is prepared to face it if it comes up, it tends to go better. 

  3. Physicians should explain to patients that there are no risk-free options, and that the medication can have adverse effects.  It is helpful to mention that adverse effects generally are reversible: they go away when the drug is stopped; and that most adverse effects can be treated easily.  Akathisia, for example, responds to treatment with beta-blockers, benzodiazepines, or Periactin.  Patients may fear that if they report adverse effects, they will be taken off the medication.  This is not necessarily the case.  They may fear that any adverse effect they have will be permanent.  This misconception could lead to hopelessness and contribute to suicide risk. 

  4. If the person prescribing antidepressant medication does not plan to see the patient weekly, appointments should be offered with the clinic nurse, if in a primary care setting; or with a psychologist or other counselor, if in a mental health clinic.  Even a scheduled phone call could help.  Doctors should set aside time in their schedules for phone calls, and/or use a secure e-mail system.  Some patients will send an update by fax to the doctor's office, with a note such as 'I just wanted you to know about this (possible adverse effect), please call me if this should be addressed.'

  5. Insurance companies should not erect bureaucratic and financial barriers to appropriate evaluation and follow-up. They should not authorize just one visit when the patient or the doctor's office calls for preauthorization of services.  They should automatically authorize two or three sessions, so that it will be more likely that proper follow-up will occur. 

  6. Accrediting agencies should monitor closeness of follow-up as an index of quality of care. 

  7. If patients have a known risk of suicide, their family and their treatment providers should ask them what would get in the way of their reaching out, should they become suicidal.  Anticipating the barriers ahead of time often makes the barriers go away. 

  8. For mild depression, consider using psychotherapy and/or lifestyle modifications before trying medication.

  9. If you do decide to take use antidepressants, use them properly.  Use a med box to help guard against skipped doses.  Do not take more than recommended.  In most cases, do not stop suddenly.  Be very careful about alcohol, nicotine, caffeine, illegal drugs, over-the counter medication, and herbal or alternative-medical products in combination with antidepressant medication.