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]


4 November 2004 

Dear Mr Medawar,

Thank you for your letters of 29 September 2004 and 7 October and your letter of 7 October 2004 to Dr Raine. I would like to take this opportunity to cover all the points you have raised in these letters. I am sorry that this reply is a few days later than I had promised.

MHRA website entry and timing
I would first like to apologise that the title of your research paper has been incorrectly quoted on the website. I would also like to offer my apologies that the differences in opinion that we obviously hold concerning the interpretation of data collected as part of the Yellow Card scheme, were not explained more appropriately on the website. I accept that the comment "The criticisms of the coding and follow up of reports are unfounded" in relation to your research could and should have been more explicit in explaining why the Agency did not agree with the conclusions you reached.

Turning now to the points you have raised in your latest correspondence, you first ask why the Agency did not accept your criticisms of Yellow Card data and why it took so long for the Agency to communicate with you. The Agency carefully assessed and reviewed the criticisms in your paper about the processes of coding and follow-up of ADR reports. The MHRA operation and procedure are scrutinised through Parliament in a variety of ways, including by the National Audit Office last year, when it specifically looked at the operation of the Scheme. The Yellow Card Scheme has subsequently been subject of an Independent Review (to which you gave evidence) and is the subject of published performance targets, reported to Parliament annually.

On the question of the time the Agency took to communicate with you, both Dr Herxheimer and you yourself gave evidence at the Independent Review of Access to the Yellow Card Scheme and you also presented evidence to the Expert SSRI Working Group. The Agency considered that the Independent Review might wish to comment on the detail of the collection, coding and interpretation of data from Yellow Cards and therefore delayed commenting on your paper until the Review’s recommendations were known. In the event the Review decided that this level of detail was outside its remit. In the light of the publication of the Review on 4 May, the Agency took the view that it was important to make reference to your paper in the context of a Question and Answer brief on topical issues relevant to Yellow Cards. Since your research paper was published, I am arranging for that entry on the to be removed.

Anonymised Single Patient Prints (ASPPs)
You ask for details of when the Agency made clear that information suppli4to you was not Anonymised Single Patient Print (ASPP) data in toto but extracted information from the ADROIT database. Together with Dr Herxheimer you attended a meeting at the Agency on 28 April 2003 when staff gave a demonstration of the ADROIT database and I understand also explained the basis of the data that was provided to you.

Following that meeting, on 5 May 2003 Dr Herxheimer wrote to Dr Raine (Annex 1) asking for additional information. Ms Wark, in her reply of 7 May 2004 (Annex 2), explained that "instead of providing ASPPs, we generally provide line listings of cases.." and a sample ASPP was also included. Ms Wark also explained that it would be impractical to provide a line listing (spreadsheet) which included every field from the ADROIT database, rather we provide fields most relevant to the enquiry. Dr Herxheimer in his later e-mail of 11 May 2003 (Annex 3) asked for details about the different fields on the sample ASPP and asked for certain specific fields (reporter source, severity of reaction, other reactions etc) and for the information contained in them to be provided in the spreadsheets.

The response from Ms Wark (Annex 4), quoted in your letter of 29 September, was directed to a specific question (no 7) raised by Dr Herxheimer in his e-mail of 11 May 2003. This answer refers to the sample ASPP given and not to the data that was provided to you and Dr Herxheimer on Excel spreadsheets.

In relation to Ms Wark’s comment on the "truncation" of text related to the data on the spreadsheets, I can accept that our reply could have been clearer, particularly as I note that Dr Herxheimer’s question (8) of 11 May used the term "ASPP". With hindsight it would have clarified the situation had the Agency taken the opportunity to explain that ‘truncation’ is not a problem with ASPPs but on the information sent on spreadsheets.

Calculated onset times
The explanation of the "onset time calculation" given in Dr Raine’s letter of 30 September was to clarify how the Agency’s database automatically calculates this. If a Yellow Card contains information about the time of onset of an adverse reaction (e.g. ‘one day after starting’), that information is recorded in the ‘free text’ area of the database and is included in any subsequent analysis. In addition, where available information associates the suspected reaction with increasing or decreasing the dose, this is also reflected in the text of the report.

Where information about the time to onset of the reaction does not correspond to the information cited in the text of the report, the database automatically calculates the time between onset from the date the drug was started and the date the reaction began. In some cases, automatic calculation of onset time may not be appropriate, for example in relation to a drug withdrawal reaction. During data entry onto ADROIT (referred to as transposition) this field is manually removed so that the information in the text reflects the report. As stated in Dr Raine’s letter of 30 September 2004, this has not occurred in all cases. However, the original Yellow Card is always reviewed along with database outputs including text when reviewing a case for drug safety signals.

Actions taken by MHRA
As you are aware, the Independent Review of the Yellow Card Scheme, has produced a large number of recommendations to strengthen the scheme and to widen access to Yellow Card data, as well as the recommendation to accept reports of suspected ADRs directly from patients. As stated in Dr Raine’s letter to you on 30 September, decisions on the remaining recommendations will be made in the light of the consultation responses. The Review also recommended the implementation of formal procedures for pre-publication submission of papers to ensure that there is adequate opportunity for MHRA to study research findings and take any regulatory action necessary.

The MHRA has contributed to a number of publications on methods used to assess drug safety signals (including use of spontaneous ADR data). As a matter of routine the Agency is continually reviewing its procedures. For example, additional guidance explaining each field extracted from the ADROIT database has been prepared to accompany ADROIT data sent out in response to an enquiry. The follow-up procedures of Yellow Card reports have been reviewed and strengthened. Agency staff have been provided with additional guidance and reminded about the importance of good data management.

Follow-up of Yellow Card Reports
In relation to our follow up requests to reporters, some do not see responding to requests from the Agency as an important priority despite the continued promotion of the Yellow Card Scheme. There is a steady expansion of those able to prescribe medicines and thus potential for widening the base of ADR reporters. We are continually stressing the importance of reporting ADRs and providing follow-up information when requested, and this is being included in the training curriculum for new groups of prescribers. In addition, patients will in due course also be able to report their experiences directly to the Agency. We therefore anticipate that there will be an increase in receipt of both initial and follow-up information.

The Yellow Cards reports detailed in section 3.3 (where the outcome was fatal) have been followed-up for further information as part of the Agency’s ongoing review of the safety of SSRIs. The judgement made on the need for individual follow up is based on the extent to which additional information would inform risk assessment decisions.

Agency procedures guide staff as to which reports should be followed up, and we have previously provided you with a copy of the standard operating procedure. Any report which contains a serious suspected reaction and is incomplete in key areas is considered for follow up. The information considered necessary to complete a report will vary depending on the nature of the suspected adverse reaction. The decision to seek additional information over and above key areas is generally left to the professional judgement of Agency staff. In some circumstances, specific drugs and reactions are selected for ‘targeted follow up’ in which case there is a minimum set of essential information which should be sought for all relevant reports. For example in the late 1990s due to concerns about withdrawal reactions associated with SSRIs particularly Seroxat, we instituted special follow up procedures. Reports of withdrawal reactions with SSRIs are routinely followed up for full details of symptoms experienced on withdrawal and of the outcome of the reaction.

Assessment of risk of suicidal behaviour and changes in SSRI concentration
The Expert Group on the Safety of SSRIs was reminded of the concerns expressed at the July 2003 meeting by both you and Dr Herxheimer about the association between changes in drug concentration and occurrence of suicidal behaviour. The Group considered the evidence supporting suicidal behaviour in temporal association with dose changes of SSRIs. The Group advised that although the available controlled data did not provide evidence to support such a link, it was plausible that suicidal behaviour may worsen around the time of dose changes, either due to changes of the course of the underlying disease or due to side effects of the drug. The Group advised that this issue would be difficult to study further in a controlled manner and that as a precaution, patients should be closely monitored for any side effects around the time of dose changes. This will be part of the Review report.

Dr Raine’s suggestion of a meeting to discuss the issues you have raised in your papers and letters, and to clarify any further questions, still stands.

Yours sincerely

Sir Alasdair Breckenridge

cc Mr Roy Alder
Dr June Raine

CH/Lct/MedawarO4l 104