I came across this little gem amongst all the old stuff I’ve got on my computer:
it’s from the House of Commons Health Select Committee – The Influence of the Pharmaceutical Industry evidence sessions – I’d just like to thank Dr Dollow from GSK for explaining this to me so well:
Q713 Dr Naysmith (Committee): We are getting a bit short of time now, but could I quickly ask GSK about Seroxat and the fact that, up until quite recently, June 2003, the company maintained that withdrawal symptoms from Seroxat were very rare, round about 0.1 per cent was the figure, and now you have to accept that it is nearly 25 per cent. How did that come about? There is evidence that you knew before June 2003 that the figure you were using was wrong.
Dr Dollow (GSK): I think it is important to put it in context. I think it is a good example of how labels change over time: so the labelling, be it in the patient information or the doctor information. I would confirm also that withdrawal effects and discontinuation symptoms have always been in the Seroxat label, so it has always been there both in the patient side and in the doctor side. It has, however, been updated over time as more information has become available to us. The fact that more information becomes available to us allows us to make a reasonable recommendation based on the data, so that we can say the frequency of those adverse events and the descriptions used for those adverse events is more appropriately labelled. Additionally, I think, when it comes to patient information leaflets, it is a very good example of how patient information leaflets can be worded more appropriately. Seroxat is an example where we have extensively done user testing. Patient information leaflets, while they legally have to be compatible with a summary of product characteristics, sometimes are written in a way which is unhelpful to patients. We have found through the Seroxat process and through extensive user testing that the patient information leaflet is now more friendly. I think we would also recommend that more user testing happen with patients, such that they are written from a blank sheet of paper whilst still being compliant with the summary of product characteristics.
Background to the question above: Until 2002, GSK stated that 0.2% of users suffered withdrawal difficulties. This increased to 0.7% and then in 2003 they admitted 25% suffered withdrawal problems. However, this information was not positively communicated to GPs, who continued to reassure their patients that Seroxat was safe and non-addictive. By the way, the current PIL (March 2007) says the figure is now 3 in 10 – that’s 30%.
The next gem, this time from the Report of the Health Committee (page 87), published 5 April 2005:
“The EWG working papers suggested that companies may not comply with requests for relevant information, and that the MHRA is often in no position to require them to. One example involved GSK, the Marketing Authorisation (MA) Holder, arguing first, that it had fully investigated Seroxat withdrawal problems, then later resisting the regulators’ proposal to warn that Seroxat withdrawal appeared particularly troublesome, on the grounds that no clinical trials had been done to establish this.
From October 2003:
The MA holder [GSK] considers that the clinical trials already conducted … have allowed the nature, frequency and severity of withdrawal reactions to be comprehensively characterised. They do not consider that further studies would add appreciably to the knowledge of events and consequently do not plan to conduct any further studies in this area.
From February 2004:
They [GSK] are of the opinion that there are no data from well designed,comparative clinical trials that would support the conclusion that the true frequency of withdrawal reactions is higher for paroxetine than other SSRIs as a class and that such statement in the SPC should not be based on spontaneous reporting data.
All the available evidence pointed to a singular risk with Seroxat, but the warnings eventually proposed by the MHRA/CSM did not mention it.”
Just what’s it going to take to get something done about this?