By Patrick Waller, Consultant in Pharmacoepidemiology, Southampton, United Kingdom.

Over the past two years, the BBC has broadcast three Panorama programs primarily focused on the safety of paroxetine. These have suggested that there are major safety issues with the drug and that our systems for detecting and dealing with such problems are seriously inadequate. The last program, aired in October 2004, raised my hackles sufficiently that I was moved to write a personal view in the BMJ (2004; 329: 865 or at http://bmj.bmjjournals.com/cgi/content/full/329/7470/865?ehom). In particular, there seemed to be a conspicuous lack of balanced coverage and an unjustified allegation was made that patients "were the last priority" of the UK regulators. It was notable that the program relied on tragic individual cases and that drug-induced causation was assumed despite the putative adverse effect (suicidal thoughts or actions) being a consequence of the principal disease indication for the drug. As those who heard several relevant presentations at the recent ICPE in Bordeaux will know, there are difficult and controversial issues here, and pharmacoepidemiology is struggling to resolve them.

Regardless of how safe paroxetine is or how effective the systems are, there are some important underlying issues here for our discipline. Perhaps the most fundamental is that few people outside the field seem to understand what it can and can't deliver. There is a substantial gap between expectations and reality. Whilst continuing to strive to provide more clearcut answers, we also need to take some time to ensure that our outputs are better understood and provide a more positive influence on public perceptions.

Television is a powerful medium and producers have little difficulty in persuading the majority of their audiences that a drug caused harm or that pharmaceutical companies and regulatory authorities are not acting in the public interest. Because much of the evidence we produce is legitimately open to considerable debate about its interpretation, experts willing to put forward particular views are not hard to find. Although the effects of scare stories relating to the use of oral contraceptives and vaccines are fairly well documented, at the moment, we can only guess what the adverse consequences of this program might be. Perhaps we should be trying to find out.

I imagine that, when writing papers to present the output of their work, most scientists spend many hours reflecting on their words and polishing them. They go through multiple drafts and seek input from others, following which their work is subjected to formal peer review. Once a paper is accepted, they carefully scrutinize the proofs. Contrast this with the situation when they have an opportunity to present their views to the public in a television documentary. There, what is being done is analogous to writing a 10,000 word paper using only a dictating machine, never seeing it written down but allowing the publisher to choose which 200 words he would like to include in the middle of an article written by someone else. And, by the way, you only get to see it on publication and, if you don't like it, you can take your complaints elsewhere. Now, who in their right mind would do that?

Of course, many of us, when asked to appear on television, agree to do so, despite the unreasonable rules of the game. I sat in front of Panorama's cameras and counted myself as lucky when they cut me out of the program altogether. After that experience I would not be keen to do so again.

There are no easy solutions. If those that are not in sympathy with the thrust of such programs refuse to appear, then imbalance is increased and it may be made to look as though they have something to hide. Pointing out that the drug may not have been responsible is perceived as uncaring if juxtaposed with bereaved relatives reliving their grief for the cameras. Arguing that a difficult balance has to be achieved (i.e. between risk and benefit) may be perceived as balancing harm to patients against the profits of the pharmaceutical industry. These are serious challenges for pharmacoepidemiology and ones we have barely addressed yet.

Return to table of contents