Reader Comments

Post a new comment on this article

Bitter pills and puffed trials

Posted by plosmedicine on 30 Mar 2009 at 23:42 GMT

Author: Stephen Senn
Position: Professor of Statistics
Institution: University of Glasgow
Submitted Date: May 22, 2005
Published Date: May 23, 2005
This comment was originally posted as a “Reader Response” on the publication date indicated above. All Reader Responses are now available as comments.

I agree with Richard Smith1 that something needs to be done about the reporting of pharmaceutical industry trials. Like him, I believe that the solution should include compulsory publication on the internet of trials 2. However, I disagree that the problem has its origin with the pharmaceutical industry; it is inherent to medical publication.

Of his eight ways of massaging data, the last five are dealt with by the International Conference on Harmonisation (ICH) guidelines covering statistical principles for clinical trials (ICH E9)3 that require pre-specification of analyses. It is not possible to claim non-inferiority on the basis of failure to prove a difference and a paper describing appropriate approaches to equivalence trials that Richard Smith thought worth publishing in the British Medical Journal4 was doing no more than explaining what was common practice within the industry. The first three points are less easily policed, although choice of control group is taken extremely seriously and, indeed, there is an appropriately entitled guideline5 that covers this.

The problems are inherent to publication not drug regulation. An instance: The New England Journal of Medicine (NEJM) published in January 2002 a paper that claimed that, Voriconazole is a suitable alternative to amphotericin B preparations for empirical antifungal therapy in patients with neutropenia and persistent fever.7 However, a letter to the editor in the same issue of that journal from scientists based at the Food and Drug Administration (FDA)8 pointed out that the analysis presented was not that pre-specified in the protocol and that not only had voriconazole failed to demonstrate non-inferiority but it was actually statistically significantly inferior to amphotericin B. Surely, responsibility for this discrepancy cannot be laid at the door of the FDA, nor can it be blamed on Pfizer. It is the authors and the NEJM that owe readers some sort of explanation. Do the editors agree with Richard Smith (and me) that a published paper, whatever else it covers, should always identify the results of pre-specified analysis and if so how do they check that this is so?

Thus, I agree with Richard Smith that much is wrong with the publication of clinical trials sponsored by the pharmaceutical industry. I disagree that it is a particular problem for industry trials. It is the publication process which is in need of reform and in particular we need to scrutinize carefully the motives of authors in publishing and the standards that editors apply in deciding what gets published.

1. Smith R. Medical journals are an extension of the marketing arm of pharmaceutical companies. PloS Medicine 2005;2(5):e138.
2. Senn SJ. Authorship of drug industry trials. Pharmaceutical Statistics 2002;1:5-7.
3. International Conference on Harmonisation. Statistical principles for clinical trials (ICH E9). Statistics in Medicine 1999;18:1905-1942.
4. Jones E, Jarvis P, Lewis JA, Ebbutt AF. Trials to assess equivalence: The importance of rigorous methods. British Medical Journal 1996;313(7048):36-39.
5. International Conference on Harmonisation. ICH harmonised tripartite guideline: choice of control group and related issues in clinical trials: International Conference on Harmonisation, 2000:1-35.
6. Hasselblad V, Kong DF. Statistical methods for comparison to placebo in active-control studies. Drug Information Journal 2001;35:435-449.
7. Walsh TJ, Pappas P, Winston DJ, Lazarus HM, Petersen F, Raffalli J, et al. Voriconazole compared with liposomal amphotericin B for empirical antifungal therapy in patients with neutropenia and persistent fever. N Engl J Med 2002;346(4):225-34.
8. Powers JH, Dixon CA, Goldberger MJ. Voriconazole versus liposomal amphotericin B in patients with neutropenia and persistent fever. N Engl J Med 2002;346(4):289-90.

Competing interests declared: The author consults frequently for the pharmaceutical industry.