Disparities Research Guidelines

Disparities Research Guidelines

These guidelines are intended to provide guidance for authors submitting works that focus on reporting health differences by race and/or ethnicity. These instructions are to encourage consistent framing, terminology, and methods aligned with established best practices for scientific work on racial and ethnic disparities in health and will be regularly updated as needed.



  • Provide a conceptual model to convey the relationships between race and/or ethnicity with other variables. Examples include but are not limited to the National Institute of Minority Health and Health Disparities Research Framework. Please reference original literature relevant to the conceptual model as applicable.
  • Social and structural forms of racism, including those related to socioeconomic status, access to care, and environmental factors should be considered and explained as applicable.



  • Be as specific as possible when acquiring information about race and ethnicity. When possible, use the 2020 U.S. Census classifications established with guidance from the Office of Management and Budget: American Indian or Alaska Native; Asian; Black or African American; Native Hawaiian or Other Pacific Islander, White; and whether of Hispanic, Latino/a/-x or Spanish origin.
  • When possible, assess and report on individuals’ racial and ethnic origins using the U.S. Census classifications that can be more specific:
    • For Asian Americans, the 2020 U.S. Census classifications are: Asian Indian; Bangladeshi; Bhutanese; Burmese; Cambodian; Chinese, except Taiwanese; Filipino; Hmong; Indonesian; Japanese; Korean; Laotian; Malaysian; Mongolian; Nepalese; Okinawan; Pakistani; Sri Lankan; Taiwanese; Thai; Vietnamese; Other Asian, specified; Other Asian, not specified; and Two or more Asian origin.
    • For Native Hawaiians and Other Pacific Islanders, the 2020 U.S. Census classifications are: Chamorro; Fijian; Marshallese; Native Hawaiian; Other Melanesian; Other Micronesian; Other Pacific Islander; Other Polynesian; Samoan; Tongan; and Two or More Native Hawaiians and/or Other Pacific Islander.
  • If race and ethnicity are codified by others, be as specific as possible when stating how the categories were attributed (e.g., assumption by medical/research personnel; algorithmic based on origin of last name).
  • Please be as clear as possible when describing the method used to categorize race and ethnicity. While some data sources may not collect these features in optimal ways; the goal is to have articles be as transparent as possible about the approach used to classify race and ethnicity.
    • Self-identification, write-in and/or ability to select multiple categories, is the ideal method to collect race and ethnicity data.
  • Be mindful to use terms related to race and ethnicity correctly:
    • Race is a heterogeneous social construct. It is correlated with environmental, social, cultural, political, and genetic ancestry.
    • Hispanic and Latino/a are ethnicities.
    • In general, use the term “White” instead of “Caucasian” because Caucasian is inaccurate.
    • In general, “Non-White” and “Other” race and/or ethnicity should not be used to capture smaller population groups. Instead, specifically identify the additional groups when possible.
  • Provide justification in the context of the scientific question when including race and/or ethnicity as a covariate in risk-adjustment models.



  • Test and report results from analyses according to the proposed conceptual framework.
  • Avoid statements of causal inference or culpability (e.g., “Black adults did not respond to x medication.” Instead, “Among Black adults, x medication was less effective.” or “Factors associated with lower effectiveness included self-identified Black race, and x”).



  • Interpret data according to the proposed conceptual model.
  • When applicable, clearly state relevant structural (e.g., racist laws; biased policies) and social (e.g., inter-racial communication; trust; social determinants of health) factors that influence the question of study.
  • In general, avoid using genetics in isolation to explain social constructs if ancestral admixture was not captured and analyzed in the statistical approach.
  • Describe implications of findings. Offer specific next steps when possible.