An empirical analysis of overlap publication in Chinese language and English research manuscripts.

Background: There are a number of sound justifications for publishing nearly identical information in Chinese and English medical journals, assuming several conditions are met. Although overlap publication is perceived as undesirable and ethically questionable in Europe and North America, it may serve an important function in some regions where English is not the native tongue. There is no empirical data on the nature and degree of overlap publication in English and Chinese language journals.

Methods/principal Findings: A random sample of 100 English manuscripts from Chinese institutions was selected from PubMed. Key words and institutions were searched in the China National Knowledge Infrastructure, a comprehensive Chinese language research database. Unacknowledged overlap was a priori defined according to International Committee of Medical Journal Editor (ICMJE) guidelines following examination by two individuals. 19% (95% CI 11-27) of English manuscripts from Chinese institutions were found to have substantial overlap with Chinese published work based on full text examination. None of the manuscripts met all of the criteria established by the ICMJE for an acknowledged overlap publication. Individual-level, journal-level, and institutional factors seem to influence overlap publication. Manuscripts associated with an institution outside of China and with more than one institution were significantly less likely to have substantial overlap (p<0.05).

Conclusions/significance: Overlap publication was common in this context, but instances of standard ICMJE notations to acknowledge this practice were rare. This research did not cite the identified overlap manuscripts with the hope that these empirical data will inform journal policy changes and structural initiatives to promote clearer policies and manuscripts.