Treatment Online

More Evidence Disputes Suicide, Antidepressant Link
> 7/6/2007 9:26:57 AM

In October of 2003, the FDA, faced with tragic anecdotal evidence and sensing a potential PR emergency, accepted the results of one study as sufficient to place a startling warning on antidepressant medications nationwide. Caution is always the safest course, but since the 2004 black box warning introduction, the number of prescriptions as well as the number of diagnoses for depression dropped sharply and many are now wondering if these warnings are in fact scaring away those who would benefit from treatment.

This month, several studies appearing in the American Journal of Psychiatry suggest that the suicide black box warning was prematurely put on antidepressants. We've written about this topic in the past, and other research has drawn the FDA's actions into question. This new evidence provides even more support for the argument that the link between suicidality and antidepressants is not a reality.

For the first of three studies appearing this month, Dr. Robert Gibbons examined over 220,000 patients in the Veterans Administration health care system. He found that veterans treated with antidepressants had fewer suicide attempts than those treated without medication. This study is important, because it looks at a very high-risk group, which is often prohibited because of the ethical implications of letting patients die for an experiment. It is true that there was no suicide protection from antidepressants found in the age group 18-24, where the FDA makes its strongest warning, but it is also true that there was no increase in suicide attempts either. It is also important to notice here that suicidal actions instead of suicidal thoughts were measured. There were no deaths at all in the pediatric FDA study, so suicidal thoughts were used instead as indicators of danger.

In a second study, Drs. Simon and Saravino focused carefully on the timeline of depression treatment. Their study plotted suicide attempts 90 days before and 180 days after the start of treatment. Connecting the dots found a clear line slopping downward from the highest point, 1 month before treatment. While the month after treatment began was still a dangerous time for patients, there were less suicide attempts than before treatment. The relatively high-risk levels lingering in the first month of treatment may have confused some psychiatrists into seeing a negative side-effect of antidepressants. The downward-sloping line is particularly convincing as evidence because it persists across age groups and even with different medications. The general effect of antidepressants was protective against suicide, and no particularly at-risk group or dangerous medication could be found.

Even if it had paid enough attention to exact timing and the month before treatment, the FDA study may not have been able to properly graph the sloping line because of flaws in the very way that it tried to measure suicide risk. Firstly, it only gathered self-reported, spontaneous reports of suicidal thoughts. This is insufficient because measurements might only be detecting a difference in the openness of patients. Perhaps only patients helped by medication felt comfortable enough to reveal suicidal thoughts. This points to a second possible flaw--confusion about what to count as dangerous suicidal ideation. It is not clear that all thoughts and discussions about suicide always increase the risk of suicide. Further work is needed to classify all the types of suicidal ideation and their respective risks. It may even be possible that the predictive value of suicidal ideation lies in a factor different altogether from the content or frequency of thoughts. For example, this study from FSU suggests that the variability in suicidal ideation is a better predictor of suicide attempts than intensity or duration. Simon and Saravino skirted this confusion by limiting measurement to clear suicidal behavior.

The FDA may finally understand the need for a standardized and rigorous classification of suicidal ideation and behavior, because it supported a new system tested by Dr. Kelly Posner, the Columbia Classification Algorithm of Suicide Assessment. The algorithm was more consistent than pharmaceutical company assessments, and it found more suicidal events but less suicide attempts when using updated criteria and search methods. Perhaps discrepancies between studies of the antidepressant-suicide link can be reconciled with a better measuring system.

The studies discussed above are just the newest pieces of a vast pile of evidence exploring the link between antidepressants and suicide; others are already in the works. With each successive completed report it looks clearer that the black box warning was rushed without thorough study. It is important to be vigilant that dangerous medication does not make it into the market, but it is also important to make sure that panic does frighten or dissuade the public from seeking beneficial and potentially life-saving medication.

Comments

No comments yet.