Potentially coercive self-citation by peer reviewers: a cross-sectional study.

Objective: Peer reviewers sometimes request that authors cite their work, either appropriately or via coercive self-citation to highlight the reviewers' work. The objective of this study was to determine in peer reviews submitted to one biomedical journal (1) the extent of peer reviewer self-citation; (2) the proportion of reviews recommending revision or acceptance versus rejection that included reviewer self-citations; and (3) the proportion of reviewer self-citations versus citations to others that included a rationale.

Methods: Peer reviews for manuscripts submitted in 2012 to the Journal of Psychosomatic Research were evaluated. Data extraction was performed independently by two investigators.

Results: There were 616 peer reviews (526 reviewers; 276 manuscripts), of which 444 recommended revision or acceptance and 172 rejection. Of 428 total citations, there were 122 peer reviewer self-citations (29%) and 306 citations to others' work (71%). Self-citations were more common in reviews recommending revision or acceptance (105 of 316 citations; 33%) versus rejection (17/112; 15%; p<0.001). The percentage of self-citations with no rationale (26 of 122; 21%) was higher than for citations to others' work (15 of 306; 5%; p<0.001).

Conclusions: Self-citation in peer reviews is common and may reflect a combination of appropriate citation to research that should be cited in published articles and inappropriate citation intended to highlight the work of the peer reviewer. Providing instructions to peer reviewers about self-citation and asking them to indicate when and why they have self-cited may help to limit self-citation to appropriate, constructive recommendations.