Description of color/race in Brazilian biomedical research.

CONTEXT AND OBJECTIVE Over recent years, the terms race and ethnicity have been used to ascertain inequities in public health. However, this use depends on the quality of the data available. This study aimed to investigate the description of color/race in Brazilian scientific journals within the field of biomedicine. DESIGN AND SETTING Descriptive study with systematic search for scientific articles in the SciELO Brazil database. METHODS A wide-ranging systematic search for original articles involving humans, published in 32 Brazilian biomedical scientific journals in the SciELO Brazil database between January and December 2008, was performed. Articles in which the race/ethnicity of the participants was identified were analyzed. RESULTS In total, 1,180 articles were analyzed. The terms for describing race or ethnicity were often ambiguous and vague. Descriptions of race or ethnicity occurred in 159 articles (13.4%), but only in 42 (26.4%) was there a description of how individuals were identified. In these, race and ethnicity were used almost interchangeably and definition was according to skin color (71.4%), ancestry (19.0%) and self-definition (9.6%). Twenty-two races or ethnicities were cited, and the most common were white (37.3%), black (19.7%), mixed (12.9%), nonwhite (8.1%) and yellow (8.1%). CONCLUSION The absence of descriptions of parameters for defining race, as well as the use of vague and ambiguous terms, may hamper and even prevent comparisons between human groups and the use of these data to ascertain inequities in healthcare.