Editor—The requirement that all authors have the idea, do all the work, get the data, analyse the data, and write the paper may be perfectly applicable to fundamental research, perhaps, or small clinical trials. In large studies it is not applicable: we are doing a 40 000 patient study of nonsteroidal anti-inflammatory drugs and COX-2 inhibitors, requested by the regulatory authorities, financed by pharmaceutical companies, driven by an independent scientific committee. Fifty people, including half a dozen statisticians, work in this study, which will generate about a hundred million bits of data. Papers will be written by medical writers under the surveillance and final approval of the scientific committee. Is this ghost writing?
May I hire a professional writer to write papers students did not or could not write, and I don't have the time to? Should these data lie ignored? Should that writer, who was not involved in the initial conception or in data collection or its analysis be an author? If not, is it ghost writing?
There is an infinity of variations between the lone searcher who does everything, and the key opinion leader who does nothing but sign.
Abbasi's simple statement that 50% of all publications are ghostwritten is misleading and derogatory,1 indicating a misunderstanding of the complexities of modern studies. It could too easily be picked up by politicians (who we all know write their speeches themselves) and others for some easy doctor bashing. There may be some cases, of course. A difficult topic, no simple answers.
Competing interests: NM must have hidden interests, certainly, although at the moment he can't recall any relevant ones.
References
- 1.Abbasi K. Editor's choice. Transparency and trust. BMJ 2004;329: 0-g. (23 October.) [Google Scholar]