REPORTING OF GENDER-RELATED INFORMATION IN CLINICAL-TRIALS OF DRUG-THERAPY FOR MYOCARDIAL-INFARCTION

Background: Concern has been expressed that women are not adequately r epresented in clinical trials evaluating treatments for medical condit ions they commonly experience. This study was designed to assess the r eporting of data on women in recently published trials of drug therapy for myocardia[ infarction, including those funded by an agency with a gender-related policy. Methods: All randomized controlled trials and meta-analyses of drug therapies for myocardial infarction published in The New England Journal of Medicine, The Lancet, The journal of the A merican Medical Association, the Annals of Internal Medicine and the B ritish Medical journal from January 1992 to December 1996 were evaluat ed. On preliminary review, 102 articles met the inclusion criteria; th ese were reviewed in detail, and 59 were excluded. Two reviewers indep endently extracted gender-related information from the 43 articles; di screpancies were resolved by consensus. Results: Women represented up to 48% of the trial participants (mean 24.1%). In the trials funded by an agency with a gender-related policy, only 16.8% of participants, o n average, were women. Of the 43 articles in the sample, only 14 (32%) provided gender-related results. Funding from an agency with a gender -related policy did not affect the reporting of gender-related informa tion. Subgroup analyses were provided for 14 (32%) of the 43 trials, i ncluding 2 (29%) of 7 trials funded by an agency with a gender-related policy. Of the 12 trials that included interaction analyses (excludin g the 2 trials in which secondary analyses were conducted specifically to identify differences between women and men), 7 (58%) conducted an interaction analysis to determine if women responded differently than men; for one of these the interaction analysis was for a secondary out come measure (drug safety). Only 5 (12%) of the 43 articles mentioned the differences between men and women in the Discussion section; 2 of these were studies that used secondary analyses to examine sex differe nces. Of the 5, only 1 was funded by an agency with a gender-related p olicy. Interpretation: Women were poorly represented in the randomized controlled trials in this sample, regardless of whether the trials we re funded by an agency with a gender-related policy. Structured report ing of gender-related information for clinical trials may improve the quality of information available about women and therefore facilitate the application of research findings to the care of women.