I am afraid there is no way I can respond to your letter of 28 January both clearly and as tactfully as I would want. Forgive me if I err on the side of clarity.

I didn't want or ask you to answer the letters I wrote to Dr Jones, so it is nothing personal if I admit to a certain ghastly fascination in charting the MCA/CSM's defensive processes in action. They seem to me to have paradoxical qualities that verge on the surreal. How can private purpose so readily serve the public interest? How can this dedication to inertia be so frenetic? How can processed words seem so bland when their meanings scream? And what of this grinding obsession with secrecy that reveals so much? When exposure to ridicule and the risk of professional disgrace seem preferable to responding to straight questions with straight answers, clearly all is not well.

For over a year, the MCA/CSM has been conducting a review that was prompted by concerns raised in The Antidepressant Web. During that time the Agency has consulted drug regulatory agencies in many other countries, all of the manufacturers and numerous clinical and academic experts. Surely the fact that your review is still in progress must, in itself, make some of your colleagues wonder if these attempts to dismiss my concerns will not prove futile. So far, the MCA/CSM has failed to come up with any substantial evidence to suggest these risks do not in fact exist, while evidence is mounting that they do.

You say that "any decision to withdraw the paper is one for the authors and not the MCA". I dispute this, since the study was based on data provided through the NHS, even if the analysis was paid for by fees from manufacturers. But perhaps we could agree that it is open to the MCA/CSM to dissociate itself from the risk assessment made?

It is all the more depressing to see the authorities so mistaken when use of the identical methodology and interpretations led yesterday's experts to suppose that with benzodiazepines there was "a dependence rate of 5-10 cases per million patient months" (Committee on Review of Medicines, 1980). This too proved a grotesque underestimate of risk.

I was not sure what you meant when you referred to the "small numbers of patients" involved in the study by Rosenbaum et al.; are you implying the numbers were inadequate? Nor is it clear what you had in mind when you said that the approach of these researchers "may overstate the clinical relevance of such symptoms". Perhaps you could explain this, because it seems to me that the US study (if not the interpretation of it) generated data that demand a complete reassessment of the MCA/CSM's view of the risks.

The US study establishes beyond reasonable doubt that depression ("acute depressive episode") is a symptom of SSRI withdrawal. The authors call it 'relapse' - though the timing clearly suggests not. But whatever it is, you would be hard put to accuse them of overstating the point - since that diagnosis was made using instruments accepted by the MCA/CSM and others, for years.

The US study also establishes beyond doubt that the numbers of patients likely to experience withdrawal symptoms, after drug exposure of only a few weeks, hugely exceeds the numbers estimated by the MCA/CSM. The US study recorded withdrawal symptoms in some 66% of patients coming off paroxetine, whereas the MCA/CSM paper suggests an incidence of 3 cases per 10,000 prescriptions. Phew! For the MCA/CSM to explain such differences in the deadpan terms used in your letter seems completely unacceptable. Would you wish to come back to me on this, or should I make it the subject of another complaint?

The point about the US study that you seem to miss completely is the striking difference between what actually goes on during the withdrawal period, and what patients and clinician seem to think is happening. The point here is that spontaneous reports focus almost entirely on somatic distress, and that 'depression' can be identified only by use of a rating scale. The clear implication - apart from the obvious risks for patients - is that the spontaneous reports on which the MCA/CSM rely may also be qualitatively very misleading in merely recording reactions "considered clinically important".

It seems to me you wilfully miss the point I was making about the inadequacy of the MCA/CSM's evidence about the long-term effectiveness of paroxetine. I'll make a formal complaint of it if you want, but it really shouldn't be necessary. My point was that the MCA/CSM approved the claim for long-term efficacy of paroxetine on the basis of a study that ignored the possibility of withdrawal reactions to paroxetine, counting all cases of apparent relapse after drug withdrawal as evidence of drug efficacy. I explained the problem with this approach in detail in my letter to the CSM Chairman of 10 June 1998 - and am still waiting for a credible reply. Given also that the study relied on was conducted jointly by an employee of the manufacturers of paroxetine and a CSM expert in receipt of company funding, your claim that the Rosenbaum et al study does not provide "any evidence" to cause you to doubt your view looks mighty thin to me.

You say you believe that current evidence does not point to a serious problem with fluoxetine - and I say that common sense indicates there is. Do you not accept that there is cross-tolerance between fluoxetine and other SSRIs?

No, I am not at all surprised you do not agree with my comments about the current UK medicines control system. I just wish you could provide some decent evidence to persuade me the MCA/CSM know what they are doing - and I bet you do too.

I am still waiting for a lot of long overdue stuff, both from you and from Mr Dunlevy. Perhaps you could now let me know when I can expect to receive it?

CLICK HERE TO READ ON

	Contents page
	List of correspondence with MCA/CSM