I was most interested to read the report by Baldwin et al (Reference Baldwin, Bobes and Stein1999), particularly as I am listed as a co-author, although I had never seen the text before and did not know it was being submitted for publication! How did such a situation arise and how can its repetition be avoided in future?

In many countries, particularly the USA, it is editorial policy to obtain consent to submission that includes all of those who participated directly in the work described; often to the extent of obtaining individual signatures. This is to be applauded. Working at the Feighner Research Institute in London, I personally treated 27 patients who were included in the above trial and informed the company concerned, on more than one occasion, that I wished to see the text of any manuscript submitted for publication prior to the event. But I received no reply.

Individual investigators should be named with their centres, rather than lumped together in an unspecified ‘study group’. It would also be helpful, in the context of any statistical conclusions reached, if the number of cases contributed from each individual centre were to be recorded in published reports.