Objective: To examine the extent to which gender bias, which has been identified as a feature of medical research internationally, is present in medical research published in South Africa.
Design: A retrospective review was undertaken of 789 articles, 106 letters and 266 editorials in 10 years of the South African Medical Journal (1986-1995).
Main outcome measures: These were gender of study subjects, proportion of women in the sample, and evidence of analysis of results according to gender of study subject.
Results: Forty-eight per cent of articles (377) and 98% of letters (104) did not mention the gender of the sample. Samples that included both genders had significantly fewer women than men, with 80% (297) of such articles and 93% (14) of such letters not presenting a comparative analysis of results.
Conclusions: These findings, similar to those of the international literature, indicate a predominant 'gender blindness' in published works. This precludes investigation of differences in the ways men and women experience disease and differential access to care.