Medical research on ethnicity continues to neglect the social context

Background: Racial and ethnic categories are socially constructed schemes that constantly evolve. Health sciences research reports often neglect to consider the social context of race and its implications. Around 2003, some scientific journals revised their guidelines when studies revealed inappropriate use. Follow-up analysis is needed to determine if researchers have recently improved reports of ethnicity.

Objective: To identify how ethnicity variables are used in research reported in JAMA following a recommendation published in May 2003.

Methods: We developed a qualitative coding instrument for ethnicity in research grounded in the literature. We coded all JAMA articles from the May and November issues between May 2004 and November 2008, yielding 46 articles over 5 years. We assessed articles for 19 characteristics, including the presence of ethnic variables, explanations of how ethnicity was categorized and assigned, justification for use, statements of risk markers and factors, inclusion of covariates like socio-economic status (SES), and more.

Results: Of the 46 articles coded, 15 reported on more than one ethnicity. Of these 15, three included a justification for including ethnicity, and 4 explained the choice of categorization. Three of the 15 included SES in the analysis and 6 included other known potential confounders such as neighborhood. None of the articles suggested genetic associations with health, but one reported ethnicity as a risk factor. Thirteen disclosed federal funding.

Discussion/Conclusion: These results are similar to a 2003 study. Scientific journal policies could promote awareness of the most appropriate use of ethnicity and how to best describe the research.