Background An honest self critique is or should be an important part of scientific output. Even though explicitly underlined in scientific authorship guidelines like CONSORT (1) and STROBE (2) we had the impression that many journal articles in rheumatology did not give this due importance. Furthermore we also had the impression that self critique was less frequent in basic than in clinical science articles.

Objectives To formally quantify self critique in clinical and basic science articles separately in 3 widely read rheumatology journals.

Methods The 2012 January, May and September issues of Ann Rheum Dis, Arthritis Rheum and Rheumatology were selected. All original, full articles with discussion sections were analyzed by 2 independent observers (FG, HY) to survey whether they were of mainly basic or clinical science content. Their discussion sections were electronically scanned to assess whether the word “limit” or its derivatives (i.e limits, limitations, etc.) (limit.der) were ever used and whether, when used, these words implied self critique. In addition each observer sought for the presence of any discussion sentences by which the authors had resorted to self critique regardless whether they had included the word “limit” or a “limit.der” A third observer (YY) had the final say in instances where the initial two observers disagreed whether an article had mainly clinical or basic science content.

Results 231 articles were surveyed. 137 (59%) were of clinical and 94 (41%) of basic science content. The Table shows the number of articles that contained the word “limit” and/or a limit.der” in the discussion sections of these articles as well as the presence of any form of self critique which included the presence of the word “limit” or a “limit. der”. The statistically significant differences between the two kinds of articles persisted when each journal was assessed separately (data not given). In 4/231(1.7%) instances the word “limit” or a “limit.der” was used in a different context.

• Download figure
• Open in new tab
• Download powerpoint

Conclusions Around one half of the original articles in our leading rheumatology journals lacked self critique. This was significantly more pronounced in basic science manuscripts. We propose that such self critique should be highlighted perhaps even more in the scientific authorship guidelines especially in those related to basic science, including animal work. Also a higher priority should be given to its presence in peer review.

References

1. Altman DG, Schulz KF, Moher D, et al, CONSORT Group. Ann Intern Med 2001;134:663–94.

2. von Elm E, Altman DG, Egger M, et al, STROBE Initiative. Lancet 2007;370:1453–7.

Disclosure of Interest H. Yazici: None Declared, F. Gogus: None Declared, F. Esen: None Declared, Y. Yazici Grant/research support from: BMS, Genentech, Centocor, Celgene, Consultant for: Abbott, BMS, Celgene, Genentech, Horizon, Janssen, Pfizer, UCB