Advice Giving in Medical Research Literature

Evidence-based health advice, i.e., clinical or policy recommendations, contributes greatly to guiding medical practice and public health policies. However, whether to give health advice, especially based on individual study results, is a controversial issue: on the one hand, such advice may lack a comprehensive review of all evidence and alternative practices; on the other hand, researchers have been encouraged to translate research findings to actionable practice. To date, limited attention has been given to understanding how and where health researchers give advice in their publications, which could be critical for assessing the quality of health advice in medical literature. In this study, we conducted a content analysis of all 4,866 sentences in the abstract and discussion sections in 100 individual study papers (both randomized controlled trials and observational studies), labeling each sentence as either “strong advice”, “weak advice”, or “no advice”. We found that most authors gave advice in individual studies, but they rarely gave advice in abstract only. The common practice is either to give advice in discussion sections only, or in both abstracts and discussions. When giving advice in both sections, authors tended to give weak and non-specific advice in abstracts, while using more sentences in the discussion sections to give strong and more specific advice, adding conditions required for the recommendations. The result suggests that most researchers support giving advice in individual studies, but they are generally cautious in giving advice in abstracts.