Factors affecting sex-related reporting in medical research: a cross-disciplinary bibliometric analysis

Background: Clinical and preclinical studies have shown that there are sex-based differences at the genetic, cellular, biochemical, and physiological levels. Despite this, numerous studies have shown poor levels of inclusion of female populations into medical research. These disparities in sex inclusion in research are further complicated by the absence of sufficient reporting and analysis by sex of study populations. Disparities in the inclusion of the sexes in medical research substantially reduce the utility of the results of such research for the entire population. The absence of sex-related reporting are problematical for the translation of research from the preclinical to clinical and applied health settings. Large-scale studies are needed to identify the extent of sex-related reporting and where disparities are more prevalent. In addition, while several studies have shown the dearth of female researchers in science, few have evaluated whether a scarcity of women in science might be related to disparities in sex inclusion and reporting. We aimed to do a cross-disciplinary analysis of the degree of sex-related reporting across the health sciences-from biomedical, to clinical, and public health research-and the role of author gender in sex-related reporting.

Methods: This bibliometric analysis analysed sex-related reporting in medical research examining more than 11·5 million papers indexed in Web of Science and PubMed between 1980 and 2016 and using sex-related Medical Subject Headings as a proxy for sex reporting. For papers that were published between 2008 and 2016 and could be matched with PubMed, we assigned a gender to first and last authors on the basis of their names, according to our gender assignment algorithm. We removed papers for which we could not determine the gender of either the first or last author. We grouped papers into three disciplinary categories (biomedical research, clinical medicine, and public health). We used descriptive statistics and regression analyses (controlling for the number of authors and representation of women in specific diseases, countries, continents, year, and specialty areas) to study associations between the gender of the authors and sex-related reporting.

Findings: Between Jan 1, 1980, and Dec 31, 2016, sex-related reporting increased from 59% to 67% in clinical medicine and from 36% to 69% in public health research. But for biomedical research, sex remains largely under-reported (31% in 2016). Papers with female first and last authors had an increased probability of reporting sex, with an odds ratio of 1·26 (95% CI 1·24 to 1·27), and sex-related reporting was associated with publications in journals with low journal impact factors. For publications in 2016, sex-related reporting of both male and female is associated with a reduction of -0·51 (95% CI -0·54 to -0·47) in journal impact factors.

Interpretation: Gender disparities in the scientific workforce and scarcity of policies on sex-related reporting at the journal and institutional level could inhibit effective research translation from bench to clinical studies. Diversification in the scientific workforce and in the research populations-from cell lines, to rodents, to humans-is essential to produce the most rigorous and effective medical research.

Funding: Canada Research Chairs.