Introduction Peer review is a volunteer process for improving the quality of publications by providing objective feedback to authors, but also presents an opportunity for reviewers to seek personal reward by requesting self-citations. Open peer review may reduce the prevalence of self-citation requests and encourage author rebuttal over accession. This study aimed to investigate the prevalence of self-citation requests and their inclusion in manuscripts in a journal with open peer review. Methods Requests for additional references to be included during peer review for articles published between 1 January 2017 and 31 December 2018 in BMC Medicine were evaluated. Data extracted included total number of self-citations requested, self-citations that were included in the final published manuscript and manuscripts that included at least one self-citation, and compared with corresponding data on independent citations. Results In total, 932 peer review reports from 373 manuscripts were analysed. At least one additional citation was requested in 25.9% (n = 241) of reports. Self-citation requests were included in 44.4% of reports requesting additional citations (11.5% of all reports). Requests for self-citation were significantly more likely than independent citations to be incorporated in the published manuscript (65.1% vs 52.1%; chi-square p = 0.003). At the manuscript level, when requested, self-citations were incorporated in 76.6% of manuscripts (n = 72; 19.3% of all manuscripts) compared with 68.5% of manuscripts with independent citation requests (n = 102; 27.3% of manuscripts). A significant interaction was observed between the presence of self-citation requests and the likelihood of any citation request being incorporated (100% incorporation in manuscripts with self-citation requests alone versus 62.7-72.2% with any independent citation request; Fisher's exact test p<0.0005).

Introduction Peer review is a volunteer process for improving the quality of publications by providing objective feedback to authors, but also presents an opportunity for reviewers to seek personal reward by requesting self-citations. Open peer review may reduce the prevalence of self-citation requests and encourage author rebuttal over accession. This study aimed to investigate the prevalence of self-citation requests and their inclusion in manuscripts in a journal with open peer review. Methods Requests for additional references to be included during peer review for articles published between 1 January 2017 and 31 December 2018 in BMC Medicine were evaluated. Data extracted included total number of self-citations requested, self-citations that were included in the final published manuscript and manuscripts that included at least one self-citation, and compared with corresponding data on independent citations. Results In total, 932 peer review reports from 373 manuscripts were analysed. At least one additional citation was requested in 25.9% (n=241) of reports. Self-citation requests were included in 44.4% of reports requesting additional citations (11.5% of all reports). Requests for self-citation were significantly more likely than independent citations to be incorporated in the published manuscript (65.1% vs 52.1%; chi-square p=0.003). At the manuscript level, when requested, self-citations were incorporated in 76.6% of manuscripts (n=72; 19.3% of all manuscripts) compared with 68.5% of manuscripts with independent citation requests (n=102; 27.3% of manuscripts). A significant interaction was observed between the presence of self-citation requests and the likelihood of any citation request being incorporated (100% incorporation in manuscripts with self-citation requests alone versus 62.7-72.2% with any independent citation request; Fisher's exact test p<0.0005). Conclusions Requests for self-citations during the peer review process are common. The transparency of open peer review may have the unexpected effect of encouraging authors to incorporate self-citation requests by disclosing peer reviewer identity.