- Neville W Goodman, consultant senior lecturera
- Accepted 10 March 1994
The criteria for authorship of the International Committee of Medical Journal Editors, quoted in the instructions to authors in the BMJ,1 are “substantial contributions to (a) conception and design, or analysis and interpretation of data; and to (b) drafting the article or revising it critically for important intellectual content; and on (c) final approval of the version to be published. Conditions (a), (b), and (c) must all be met.” In an American study of 200 papers published in or before 1989 one quarter of authors did not contribute substantially.2
Methods and results
I sent a questionnaire to the first authors of all research papers that had three or seven or more authors and were published in five consecutive issues of a peer reviewed general medical journal in 1993. The questionnaire listed 16 types of contribution towards setting up a study and submitting the results for publication without indicating their importance in satisfying the international criteria for authorship; it asked the first author to tick what each of the coauthors had contributed and assured confidentiality. The table shows the contributions and whether they fulfil the criteria.
Twelve out of 14 questionnaires were returned. Only two first authors indicated that they were not concerned about confidentiality. The 12 papers had 92 authors. I excluded all but the first author on one paper with nine authors because they were all listed as having made almost all the contributions. Of 84 authors, therefore, 32 fulfilled the criteria for authorship and 19 possibly did so (51, 61% (95% confidence interval 50% to 71%)). After I had excluded another paper on a large multicentre trial 44 out of 69 authors satisfied possible and definite criteria for authorship (64% (52% to 75%)).
For the 84 authors, the median number of contributions attributed to first authors was 10 (range 5-13), to second authors 3 (1-10), to third authors 3 (1-7), and to subsequent authors (excluding the last) 2.5 (1-6). Last authors scored 4 (2-6). The final version was approved by all authors in only five papers. Six heads of department were authors without fulfilling any of the definite criteria.
About one third of authors in this small survey had not made “substantial contributions” to the intellectual content of the papers. This fraction might have been larger if the possible criteria had been more specific—for example, analysing data may just have been simple manipulation on a computer. I cannot comment on the validity of the responses except for the paper I excluded because all authors had been listed as making nearly all the contributions, but I did promise confidentiality.
Those who win grants, head departments, refer patients, measure variables, and apply standard statistical tests are important in science, but they should receive credit for what they have done and no more.3 A recent editorial asked if academic institutions are corrupt.4 An institution cannot be corrupt; only people can be corrupt. But the way an institution works can be corrupting. The current lax view of authorship is corrupting, and it is “a fiction that authorship is synonymous with authorship listings.”4
The results of this small survey on papers published in 1993 are much the same as those of the American study (published after my data had been analysed)2; authors seem no more aware of conditions for authorship now than four years ago. Journals should ask authors to fill in a questionnaire similar to the one I used and provide a published table of contributions to the paper.5