Article type
Year
Abstract
Background: the lack of consideration of sex and gender in research reduces the applicability of findings and jeopardises its capacity to support decisions. Studies suggest that such biases are related to the underrepresentation of women in the scientific community. In 2014 Cochrane’s Sex/Gender Methods Group (SGMG) published guidance to address the issue of sex/gender for systematic reviews of HIV/AIDS and hypertension.
Objectives: we described the impact of the SGMG recommendations on sex/gender reporting and analysis in Cochrane systematic reviews (SR) of interventions for HIV/AIDS and hypertension. In addition, we evaluated the association between sex/gender consideration and gender of authorships.
Methods: pre-post uncontrolled retrospective study. We screened SR of interventions for HIV/AIDS and hypertension published between 1995 and 2018. We removed reviews addressing sex-specific medical conditions. We classified SR as pre- or post-guidance of SGMG (title registered in Archie after 1 January 2015). We collected data on gender of first and last authors. We documented the frequency of sex/gender terms used in each section of the SR. In the results section, we separated descriptive information from primary studies and analytic approaches. We used ‘not applicable’ to denote a situation where insufficient primary studies or data on estimates did not enable us to conduct the intended analyses. We performed descriptive statistics, comparative statistical tests and regression analyses.
Results: of the 191 SR screened, we excluded 25 due to sex-specific conditions. We included 96 SR for HIV/AIDS (84% pre-guidance) and 70 SR for hypertension (63% pre-guidance). Women represented 53% (n = 88) and 25% (n = 40) of first and last authorship respectively. In 14% of SR both authors were women. Overall, hypertension and HIV/AIDS reviews’ reporting of sex/gender per sections are shown in Table 1 and Table 2, respectively. Logistic regression showed that reporting sex was three-fold more likely in the hypertension group (odds ratio 3.42, 95% confidence interval (CI) 1.30 to 8.98; P < 0.05) adjusted by gender, number of authors and publication of the reporting guidance. Studies with female first-last authorship had a non-significant effect on reporting sex (risk ratio 0.77, 95% CI 0.57 to 1.05; P < 0.05).
Conclusions: consideration of sex/gender in Cochrane SR for HIV/AIDS and hypertension is scarce. The guidance implementation only showed significant increase in reporting in the methods section. The differences found between the SR of HIV/AIDS and hypertension were probably due to the pathology, rather than to the effect of the SGMG guidance. Studies with a female first or last author did not show an increased probability of reporting on sex.
Patient or healthcare consumer involvement: there were no patients or healthcare consumers involved in this project.
Objectives: we described the impact of the SGMG recommendations on sex/gender reporting and analysis in Cochrane systematic reviews (SR) of interventions for HIV/AIDS and hypertension. In addition, we evaluated the association between sex/gender consideration and gender of authorships.
Methods: pre-post uncontrolled retrospective study. We screened SR of interventions for HIV/AIDS and hypertension published between 1995 and 2018. We removed reviews addressing sex-specific medical conditions. We classified SR as pre- or post-guidance of SGMG (title registered in Archie after 1 January 2015). We collected data on gender of first and last authors. We documented the frequency of sex/gender terms used in each section of the SR. In the results section, we separated descriptive information from primary studies and analytic approaches. We used ‘not applicable’ to denote a situation where insufficient primary studies or data on estimates did not enable us to conduct the intended analyses. We performed descriptive statistics, comparative statistical tests and regression analyses.
Results: of the 191 SR screened, we excluded 25 due to sex-specific conditions. We included 96 SR for HIV/AIDS (84% pre-guidance) and 70 SR for hypertension (63% pre-guidance). Women represented 53% (n = 88) and 25% (n = 40) of first and last authorship respectively. In 14% of SR both authors were women. Overall, hypertension and HIV/AIDS reviews’ reporting of sex/gender per sections are shown in Table 1 and Table 2, respectively. Logistic regression showed that reporting sex was three-fold more likely in the hypertension group (odds ratio 3.42, 95% confidence interval (CI) 1.30 to 8.98; P < 0.05) adjusted by gender, number of authors and publication of the reporting guidance. Studies with female first-last authorship had a non-significant effect on reporting sex (risk ratio 0.77, 95% CI 0.57 to 1.05; P < 0.05).
Conclusions: consideration of sex/gender in Cochrane SR for HIV/AIDS and hypertension is scarce. The guidance implementation only showed significant increase in reporting in the methods section. The differences found between the SR of HIV/AIDS and hypertension were probably due to the pathology, rather than to the effect of the SGMG guidance. Studies with a female first or last author did not show an increased probability of reporting on sex.
Patient or healthcare consumer involvement: there were no patients or healthcare consumers involved in this project.
PDF