Dear Dr Jones,

Many thanks for your letter of July 26th. Unfortunately for all of us, your reply raises more questions than it answers. From what you have sent, it seems to me that you are passing on summaries of their data presented to you by the companies. I have no doubt that you are doing this in good faith but this is where the problems start.

I am sitting here with a study in front of me that is incompatible with the summary on sertraline you have presented. I am bound by a confidentiality agreement not to divulge this to anyone. But from a legal deposition I have recently been involved in, which could be forwarded to you, and from prior discussions with the study investigator, I can tell you that this was conducted by Ian Hindmarch. It was conducted before suicide on SSRIs became an issue. The significance of this is that no-one would have had cause to explicitly code for suicidality. However for a company to say to you regarding this study that there was no suicidality would be misleading. Suicidality may not have been reported but this is clearly a different state of affairs.

This latter point applies to the Hindmarch and no doubt to many other studies. In the Hindmarch study all subjects randomised to sertraline dropped out within days with reactions that were clearly of a psychiatric nature. The fact that all dropped out and did so with the problems reported on the original form is incompatible with a description of the problems as mild to moderate.

The descriptions of these reactions that I have in front of me, combined with our healthy volunteer study, would I am certain give all ethical committees and the insurers of healthy volunteer studies a problem sanctioning studies of this type. They might permit a study to go ahead but only with warnings and close monitoring. I am fairly certain that you would make a similar judgement were you an ethics committee chairman or were you working for an insurer. Given this, I can only believe that the MCA have not received the documents am privy to. And if you have not received this document, in how many cases have you received copies of the original studies?

Where do we go from here? I feel I have little option but to draw attention to the situation by whatever means I can - the situation being that mature medical and nursing people would not be let take these drugs without detailed warnings and monitoring while children and an ever larger number of healthy individuals with stress reactions of one sort or another are being given these agents without any warnings or monitoring. Your forthcoming advice in Current Problems in Pharmacovigilance, it seems to me, will do nothing to help the situation and indeed may even aggravate it. The idea that patients thought to be at risk should be carefully monitored implies that patients already seriously suicidal may be made worse and those who are less obviously depressed are not at risk. In fact, I think the evidence all points the other way. It is those a GP might think were least at risk who in fact are at the greatest risk.

As regards your contention that there are a number of epidemiological studies that have failed to find a risk, this is clearly not the case. The only epidemiological study of any substance was done by Jick and colleagues and this produced an extremely worrying finding. If there are other epidemiological studies perhaps you could refer me to them.

As regards proper analyses of clinical trial data, these have not failed to find an association. There have been 2 analyses. One conducted by Pierre Fabre, finding a greatly increased rate of suicidality on SSRIs. The other by Beasley is so seriously flawed that it provides grounds for concern if you or the CSM are depending on this (see enclosed). My analysis of other company clinical trial databases gives a significant increase in suicide and suicide attempt rates on a range of SSRIs. An article by Khan in the Archives of General Psychiatry this April also shows an elevation in risk as does an as yet unpublished analysis by Ross Baldessarini of Harvard. In many of these analyses the SSRIs show up as having an elevated risk compared to placebo where the older or other antidepressants have a lower rate than placebo.

Perhaps you would let me know where you propose to go from here. At the very least you might let me know whether I am incorrect to assume you are depending on summaries provided by the companies and that you have not had sight of the original Hindmarch study. . Unless I hear from you to the contrary, I will report my understanding of the situation as I have outlined it and leave others to decide whether this is appropriate or not.

David Healy MD FRCPsych

Director, North Wales Department of Psychological Medicine

 CLICK HERE TO READ ON

	Contents page
	What's New?