MHRA-LOGO.gif (6825 bytes) Medicines and Healthcare products
Regulatory Agency

Market Towers
1 Nine Elms Lane, London SW8 5NQ


Mr Charles Medawar
Social Audit Ltd
P.O. Box 111
London NW1 8XG

General enquiries
Telephone 020 7273 0000  Fax 020 7273 0353
E-mail [email protected]

Direct line:  020 7084 2534
Direct fax:  020 7084 2548

Email:   [email protected]


14 January 2005

Dear Mr Medawar,

I am sorry that it has not been possible to send you an earlier reply to your letter of 6 November requesting copies of documentation relating to the Agency’s paper before the Agency’s statement (that the criticisms contained in your report concerning the coding and follow up of reports were unfounded) was published

The Agency’s view arose, in part, from discussions held at the Expert Working Group and from meetings between yourself and the Agency about the database itself. I enclose copies of the relevant extract of minutes of the SSRI Expert Working Group of 22 July and 25 November 2003 (see Annex A) and the agreed record of the evidence that you and Dr Herxheimer gave to the Independent Review of Access to Yellow Card Data, which is also attached (see Annex B).

However, as I explained in my earlier letter the statement posted by the Agency could, and should, have been more explicit in explaining why the Agency did not agree with conclusions you reached. I accept that the wording of the comment as it appeared could have been expressed more explicitly and I repeat my apologies for that

Yours sincerely
Sir Alasdair Breckenridge








(Paper provided)

5.1 The Chairman welcomed Professor Herxheimer and Mr Medawar. The Chairman, members of the Group and members of staff of the MHRA introduced themselves

5.2 Professor Herxheimer began the presentation and informed the Group that their presentation would largely follow the discussion paper which had been circulated to Group members previously (presented at Annex 1). Professor Herxheimer informed the meeting that he and Mr Medawar had written a paper on the basis of data received from the MHRA Yellow Card Database in relation to suicidal behaviour. A detailed analysis of these data had been submitted to a journal although the publication date was unknown. A shorter version had been submitted to the BMJ and although it had been initially rejected, it had been resubmitted. The conclusion of their analyses of these data and of the Panorama/Mind data was that patient reporting added value the consideration of the issues.

5.3 Suicidal behaviour

5.3.1 Mr Medawar and Professor Herxheimer informed the Group of their theory that behavioural adverse effects were associated with rapid changes in concentration of the SSRI during initiation of therapy, change of dose or withdrawal of medicine. They informed the group that they considered that some patients were being given too high a dose. They considered that this was partly because the licensed dose of was inappropriate for some patients and did not allow for titration because of lack of alternative dose forms. They further expressed their view that the liquid formulation available for some SSRIs, was difficult to use. They stated that 8% of the population were slow metabolisers and that this was the vulnerable patient group, although currently it was not possible to routinely identify these people. They expressed their view that evidence from the Yellow Card data suggested that a prescribers response to reports from the patient of suicidal thoughts was generally to increase the dose which in their view was not always the appropriate action.

5.3.2 The Group questioned whether the analysis of the Yellow Card data was affected by reporting bias given that an event was possibly more likely to be linked to a medicine by the reporter if it was temporally associated with either initiation of therapy or a dose change. Both Professor Herxheimer and Mr Medawar agreed that this was conceivable but not possible to confirm on the data available.

5.3.3 The Group commented that it would not be possible to determine whether the temporal relationship between the dose increase and the ADR could signify a change in severity of illness or a deterioration in mental health which could explain the occurrence of suicidal behaviour.

5.4 Withdrawal reactions

5.4.1 Mr Medawar and Professor Herxheimer expressed their view that the paper by Price et al (1996) on withdrawal reactions, which was an analysis of Yellow Card data, should be revised. On the definition of ‘Drug dependence’ they expressed the view that arguments as to the definition of drug dependence had not been helpful the past in communicating the risk of withdrawal reactions with SSRIs

5.4.2 Mr Medawar and Professor Herxheimer expressed their view that the long term effectiveness of SSRIs had not been demonstrated and that in some trials symptomatology of withdrawal reactions was seen as relapse and evidence drug was not working, which in turn led to increased dosages and longer duration of treatment.

5.5 Product information

5.5.1 Mr Medawar and Professor Herxheimer informed the Group that the considered that changes to the paroxetine product information were required. They commented that in their view the recent addition of frequencies of reactions on stopping paroxetine and placebo implied that the severity was the same on stopping paroxetine as placebo which they considered inappropriate. They commented that product information should be the same for all SSRIs with any differences being on the basis of clear evidence and that a precautionary principle should be applied when consideration was being given to warnings in SPCs. They further commented that information should be adapted for different indications/types of user.

5.5.2 Mr Medawar and Professor Herxheimer expressed the view that past misunderstanding, which they considered had been brought about by differentiating withdrawal reactions from dependence, should be explained. They stated that there was danger in over-emphasising concern about the effect of warnings information on compliance in severely depressed patients, as, in their view, most patients taking these medicines are not severely depressed.

5.5.3 The Group commented that it was a challenge to disseminate the information in such a way that informed patients of the risks but did not stop them taking the medication.

5.5.4 Professor Herxheimer agreed that this was a challenge but expressed the view that it was important that people should make fully informed choices for treatment and that tools such as flow charts and education materials should be used to allow the doctor and patient to work together and discuss the uncertainties of the data.

5.6 Wider communication

5.6.1 Mr Medawar and Professor Herxheimer informed the group that they considered that good communications are essential to rebuild trust between regulators, the medical profession and the public and that the Group should consider interim advice in advance of their final report.

5.6.2 The Group asked what Professor Herxheimer and Mr Medawar considered the interim advice from the Group should be. The Chairman noted that this was a difficult question and that the Group would be happy to receive an answer by letter after the meeting. Mr Medawar expressed willingness to comment on any draft wording proposed by the group.

5.6.3 Professor Herxheimer informed the Group that he considered that patients should be told that it would be difficult to stop medication and analogies to nicotine addiction might be appropriate. He also stated that advice should be provided that paroxetine and venlafaxine may be worse than other SSRIs and that patients may want to switch to a medicine with a longer half-life. He commented that in his opinion there was insufficient research on how to discontinue SSRIs.

5.6.4 The Group asked Mr Medawar and Professor Herxheimer if they had any insight as to why the majority of patients tolerate SSRIs well and why others do not.

5.6.5 Professor Herxheimer considered that this may be due to genetic differences. Mr Medawar compared this situation with that with benzodiazepines and expressed the view that it was due to differences in personality and social circumstances although there was no strong evidence to back this up.

5.6.6 The Chairman thanked Professor Herxheimer and Mr Medawar for their presentation and for answering the questions from the Group".








"6.4 Further evidence from Mr Medawar and Dr Herxheimer

6.4.1 The Group was provided with a copy of the publication from Mr Medawar and Dr Herxheimer (Medawar C and Herxheimer A. International Journal of Risk & Safety in Medicine 2003; 16:5-19). This publication compared reports of withdrawal reactions and suicide and suicidal ideation in association with paroxetine that have been received through the Yellow Card scheme from healthcare professionals with similar reports from users collated by Mind/Panorama.


The authors expressed concerns about the miscoding of reactions reported through the Yellow Card Scheme and about proper follow up of reports. The Group was informed that Mr Medawar and Dr Herxheimer had presented this paper and their conclusions to the Review of Access to Yellow Card Data and their criticisms of the Yellow Card Scheme would be taken into account in the context of the review."






Committee members

Dr Jeremy Metters (Chairman)

Professor Christine Bond

Ms Kristin McCarthy


Mr Charles Medawar

Dr Andrew Herxheimer


Dr June Raine

Mr Jeremy Mean

Ms Amanda Lawrence




1. The Chairman thanked Mr Charles Medawar and Dr Andrew Herxheimer for attending and set out the background to the Review. The Chairman Committee stated that the Committee wanted to hear their views on access to Yellow Card data and that it would be important to have a free ranging discussion.


2. Mr Medawar and Dr Herxheimer stated that they had brought a joint, written response to the consultation exercise, which specifically answered the questions set out in the public consultation letter.

3. Mr Medawar and Dr Herxheimer explained that they had both been involved in the Panorama programme on the SSRI Seroxat and that they had asked the Medicines and Healthcare products Regulatory Agency (MHRA) for copies of the anonymised patient prints (ASPPs), derived from Yellow Card reports, on withdrawal and suicidal behaviour linked with Seroxat. They stated that the data provided had been analysed, and the results were soon to be published in the International Journal of Risk and Safety in Medicine (annex 1) and that the main findings had also been presented to the Committee on Safety of Medicines (CSM) Working Group on the safety of SSRIs. In their view there were weaknesses in the completeness of the coding of reports and in the follow up of reports.

4. Mr Medawar and Dr Herxheimer shared the view that everyone in the scientific community should have access to Yellow Card data, to increase the range of expertise and the capacity to make the fullest use of the data. In terms of confidentiality their shared view was that this amounted to the privacy of the patient and that everything else was scientific data and should be more widely available.

Patient reporting

5. Dr Herxheimer and Mr Medawar stated that they were in favour of patient reports but that they should not be fitted into the existing system and that such reports would be useful for more intensive analysis.

6. Mr Medawar and Dr Herxheimer suggested that Yellow Cards should be two parts, one for the health professional to complete and one for the patient. Both of the parts should be marked with a reference number so they could be matched up after being submitted to the MHRA.

7. Mr Medawar and Dr Herxheimer did not agree that patient reporting should be through NHS Direct as they thought the suspected adverse drug reaction (ADR) would be translated from patient language into the nurse’s terminology before reaching the Agency, which would result in a ‘Chinese whispers’ effect. They also thought that complaints from patients about ADRs should be treated as Yellow Card reports. They stressed that there were almost no data on how reports of suspected ADRs from patients could be satisfactorily encouraged, collected and assessed, and that therefore a variety of methods to do this needed to be tried and critically evaluated. The DoH should consider funding several pilot projects to do this, and invite a range of potentially interested organisations to submit proposals.

Possibilities for improving the Scheme and increasing access and openness

8. Mr Medawar and Dr Herxheimer stated that one possibility would be to put anonymised ADR data on to the Internet, as they suggested had been done in Denmark. They considered that this would encourage much better standards of reporting and improve communication of information on possible ADRs.

9. The Chairman asked how we could improve on deficient and less valuable reports. Dr Herxheimer and Mr Medawar stated that although a great deal of reports concerned well known, non-serious reactions, we did not want to risk discouraging reporting. They considered that although the Agency was particularly interested in serious reactions and reactions to drugs on the black triangle list, it was important that the Agency was still made aware of other less serious reactions.

10. The Chairman asked how Mr Medawar and Dr Herxheimer thought the data should be shared. They stated that researchers should be allowed to look at different sections of the data as they had done with the data on Seroxat. They stated that greater openness would enable audit, facilitate patient involvement and feedback to reporters and patients.

11. Mr Medawar and Dr Herxheimer suggested that more work should be done with organisations responsible for undergraduate and postgraduate training to raise the level of knowledge about drug safety and to encourage reporting.


12. The Chairman asked if it was appropriate to pay a fee to reporters for submitting data to the Yellow Card Scheme. Dr Herxheimer and Mr Medawar suggested that it would be reasonable for a fee to be paid for follow up information but not for the initial Yellow Card submitted.


13. In summary Mr Medawar and Dr Herxheimer thought that anonymised data should be released as a matter of course as long as the rights of personal privacy were not compromised. They stated that while their view was that the system was imperfect, the Yellow Card Scheme was essential for drug safety.