Objectives: To identify and understand, through data from multiple sources, some of the factors that affect authors' and editors' decisions to use reporting guidelines in the publication of health research.
Design: Mixed methods study comprising an online survey and semi-structured interviews with a sample of authors (online survey: n = 56; response rate = 32%; semi-structured interviews: n = 5) and journal editors (online survey: n = 43; response rate = 27%; semi-structured interviews: n = 6) involved in publishing health and medical research. Participants were recruited from an earlier study examining the effectiveness of the TREND reporting guideline.
Results: Four types of factors interacted to affect authors' and editors' likelihood of reporting guideline use; individual (e.g., having multiple reasons for use of reporting guidelines); the professional culture in which people work; environmental (e.g., policies of journals); and, practical (e.g., having time to use reporting guidelines). Having multiple reasons for using reporting guidelines was a particularly salient factor in facilitating reporting guidelines use for both groups of participants.
Conclusions: Improving the completeness and consistency of reporting of research studies is critical to the integrity and synthesis of health research. The use of reporting guidelines offers one potentially efficient and effective means for achieving this, but decisions to use (or not use) reporting guidelines take many factors into account. These findings could be used to inform future studies that might, for example, test the factors that we have identified within a wider theoretical framework for understanding changes in professional practices. The use of reporting guidelines by senior professionals appears to shape the expectations of what constitutes best practice and can be assimilated into the culture of a field or discipline. Without evidence of effectiveness of reporting guidelines, and sustained, multifaceted efforts to improve reporting, little progress seems likely to be made.