Department of Health
Market Towers 1 Nine Elms Lane London SW8 5NQ
Telephone 0171-273 0729
Facsimile 0171- 273 0286 .
2 March 1999

Dear Mr Medawar,

Further to my letter of 25 February, I am replying to the points raised in your letter of 4 February to Mr Alder on which I have sought colleagues' advice.

I note the general points you make about MCA procedures on the question of SSRI risk assessment. Our assessment of the available data has not led us to conclude that SSRIs are drugs of dependence. As I believe you are aware, this issue is now being considered at a European level We have undertaken to keep you informed of developments and will let you know if our assessment of the evidence changes.

We have provided our views on the different methodologies used in the studies by Price et al and Rosenbaum et al on which you take issue and we can add nothing further. We do not have a reliable estimate of the frequency of withdrawal reactions associated with SSRIs. In fact, I understand that studies to investigate the issue have resulted in widely differing estimates probably due to the different study methodologies.

You ask for clarification of our view of the Rosenbaum study. We consider that it is inappropriate to compare the results of a clinical trial where a selected, tightly controlled population of patients is closely observed for, and questioned about, symptoms with the results of analysis of spontaneous reporting where many factors affect what is reported.

You state that the US study shows evidence of withdrawal symptoms much greater than MCA/CSM state. Our position regarding the frequency of withdrawal reactions is presented above.

You question the MCA reliance on spontaneous reporting. I understand that the limitations of spontaneous reporting are well known and that they were taken into account when these data were analysed. Many other sources of data are used in the assessment of drug safety issues including those used in the MCA assessment of SSRIs presented to the CSM last March, a copy of which we have sent you.

We consider that the current evidence does not indicate a serious problem with fluoxetine. I am afraid we do not understand your point about 'cross-tolerance' between fluoxetine and other SSRIs.

I hope this reply helps clarify our position. Clearly, our assessment of the scientific evidence differs from yours. In this respect, we will have to continue to disagree. We will of course continue to deal with requests for information, but cannot enter into further extended debate on this matter.

Yours sincerely
Peter Dunlevy
Executive Support


Contents page
List of correspondence with MCA/CSM