Good practice in publication of clinical trial results

As the name implies, ghostwriting is often hard to detect, so Healy & Cattell (2003) have made a valuable contribution to our understanding of this important subject by their measurement and thoughtful analysis of the practice. It is also refreshing to see such a balanced account which discusses both the benefits and potential dangers of ghostwriting.

Perhaps their most alarming observation is that the papers sponsored by the manufacturer reported ‘universally positive results’, which implies the existence of considerable publication bias. Such distortions to the published literature probably exist across all therapeutic areas and have been shown to distort the outcomes of meta-analyses (Tramèr et al, 1997) and therefore to have serious implications for evidence-based medicine (Melander et al, 2003).

Readers may be interested to know that guidelines have recently been published which call on pharmaceutical companies to endeavour to publish results of all clinical trials of marketed products (Wager et al, 2003). The guidelines also provide recommendations to ensure that professional medical (ghost) writers are used appropriately so that their contribution can be beneficial rather than harmful. The Good Publication Practice (GPP) for pharmaceutical companies guidelines have been publicly endorsed by several drug companies and communications agencies. Further details are available at http://www.gpp-guidelines.org.