There was a sneaky revision of the paroxetine/Seroxat SPC ("Summary of Product Characteristics") on 13 September – the sixth change in the last three years. Haven’t checked, but this must be some sort of record for a drug that’s been on the market for 15 years. Needless to say, the overwhelming majority of Seroxat prescribers will be unaware of the changes; they’d need to tackle the small print with a nit comb to take it all in.

The major change seems to be a very belated contraindication against concurrent use of paroxetine and pimozide. Why did Canadian doctors get a Dear Doctor letter about the same problem, more than a year ago? An FOI request has been sent to the MHRA.

If you look hard enough, you’ll also find some glaring inconsistencies, plus a number of niggling changes in the text of the new SPC – many of them softening warnings previously given. One notable example qualifies the advice that "paroxetine should not be used in the treatment of children and adolescents". The new SPC has dropped the statement that "efficacy has not been adequately demonstrated", replacing it with a sop for prescribers who decide to treat anyway: "If based on clinical needs, a decision to treat is nevertheless taken, the patient should be carefully monitored for the appearance of suicidal symptoms …" How on earth are prescribers expected to keep up to date with stuff like this?

But for all these niggles, the three main specific defects in the SPC have been there for years:

  1. In common with many other SPCs, there is routine denial of the underestimation of the incidence of different adverse effects – with calculations based essentially on the NERO fallacy, i.e. the supposition that, 'no evidence of risk = evidence of no risk'.
  2. No distinction is made between severity of adverse drug effects recorded on either placebo or active drugs – e.g. the Seroxat/paroxetine SPC claims that the incidence of paroxetine drug withdrawal reactions is only 10% (‘common’) rather than 30% (‘very common’), because of a 20% incidence of "withdrawal reactions" found with placebo. This seems both pernicious and absurd.
  3. There is still no warning in the SPC of lack of evidence of SSRI and paroxetine efficacy in cases of mild depression – though these account for two out of every three cases treated. In short, most people prescribed an SSRI have no basis for expecting real benefit, but face significant risks of harm.

The SPC prescribing advice system is a mess, and demonstrates the continuing incompetence of the MHRA. As the Parliamentary Health Committee observed (2005), the medicines control system is in dire need of review:

"During this long inquiry we became aware of serious weaknesses in the MHRA. Worryingly, in both its written and oral evidence the Agency seemed oblivious to the critical views of outsiders and unable to accept that it had any obvious shortcomings, except those that could be remedied by more transparency. The Agency’s attitude to its public health responsibilities suggested some complacency and a lack of requisite competency, reducing our confidence in its ability to undertake the reforms needed to earn and deserve public trust. Nor did we conclude that the MHRA provides the discipline and leadership that this powerful industry needs. We recommend that there be an independent review of the MHRA …"

Charles Medawar
26 September 2006

Declaration of interest:  CM was  engaged as one of four specialist advisers to the House of Commons Health Committee, on its recent enquiry into "The Influence of the Pharmaceutical Industry".