Table 2.

Detailed Information on Use of the Discrimination in Health Care Measure

Study	Discrimination Basis	Health Care Setting	Number of Items & Response Scale	Construction	Psychometric Properties	Statistically Significant Results from Multivariable Analyses	Non-Significant Results from Multivariable Analyses
Original Studies
(Bird & Bogart 2001)	Race- or color-based; or SES- or social class-based discrimination.	General health care settings.	7 race-based items & 7 SES-based items rated on a binary scale (Yes/No).	For each item set (i.e., race-based or SES-based), scores were summed then a measure of whether participants had ‘none’ versus ‘any’ of the experiences was created.	Not reported.	None reported.	None reported.
(Bird & Bogart, 2003)	Race- or color-based discrimination.	Health care settings where family planning or birth control services accessed.	8 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always).	Responses were recoded as binary (i.e., ‘never’ versus ‘ever’) for each item, then summed, with higher scores indicating greater discrimination.	Item responses were recoded as binary (‘never’ versus ‘ever’) Kuder-Richardson 20 = 0.88	None reported.	None reported.
(Bird et al., 2004)	Race- or color-based; or SES, position, or social class-based discrimination.	Health care settings where HIV treatment accessed.	7 race-based items & 7 SES-based items rated on a 5-point scale (Always, Most of the Time, Sometimes, Rarely, Never).	Responses were recoded as binary (i.e., ‘never’ versus ‘ever’) for each item, then summed for all 14 items and for each item set. In addition, for each item set, mean score was computed, with higher scores indicating greater discrimination.	Cronbach’s alpha for race-based discrimination items = 0.92 Cronbach’s alpha for SES-based discrimination items = 0.95	None reported.	None reported.
(Thorburn & Bogart, 2005)	Race- or color-based discrimination.	Health care settings where family planning or birth control services accessed.	9 items rated on a 5-point scale (Always, Most of the Time, Sometimes, Rarely, Never).	Two subscales were created based on the exploratory factor analysis results. Subscale scores were recoded as “no discrimination” (i.e., scores of 0 across items) and “some discrimination” (i.e., scores of 1 or greater).	Item responses were recoded as binary (i.e., ‘never’ versus ‘ever’) Exploratory factor analysis results supported a two-factor solution. The first factor consisted of the first four items and captured more general experiences of health care discrimination (Kuder-Richardson 20 = 0.89). The second factor consisted of four items and reflected stereotypes about African Americans (Kuder-Richardson 20 = 0.80).	Stronger black identity, younger age, and lower income were associated with reports of race-based discrimination when accessing family planning or birth control services.	Hispanic ethnicity, education level, religiosity, if the participant was currently working, number of people income supports, marital status, if the participant was currently living with a partner, number of live births, if the participant ever had an STD, if the participant had ever been testing for HIV, if the participant was currently using birth control, if the participant had vaginal sexual intercourse in the past 3 months, or number of vaginal sexual partners in the past 3 months were not significantly associated with race-based discrimination when accessing family planning or birth control services.
Found Articles
(Bogart, 2001)	Race- or color-based discrimination.	General health care settings.	7 items rated on a binary scale (Yes/No).	The number of experiences (items) endorsed was summed.	Not reported	None reported.	None reported.
(Burgess et al., 2014)	Race- or color-based discrimination.	General healthcare settings.	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always).	Mean score across items was computed, with higher scores indicating greater discrimination.	Cronbach’s alpha = 0.89	None reported.	There were no statistically significant differences between the self-affirmation intervention and control groups on discrimination in health care. Because of randomization, only bivariate analyses were performed.
(Gonzales et al., 2013)	Discrimination based on being Native American/American Indian.a	General health care settings.	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always).	A binary measure was created to categorize responses as ‘never’ (to all items) versus ‘any’ (rarely or more to at least one item). Mean score across items was also computed, with higher scores indicating greater discrimination.	Kaiser criterion (for a one-factor solution) = 0.74 Standardized item factor loadings ranged from 0.77 to 0.90 Cronbach’s alpha = 0.94	Discrimination in health care was associated with significantly higher adjusted odds of not being current for clinical breast examination and PAP test. Higher mean levels of discrimination in health care were also associated with a higher number of suboptimal health care seeking behaviors (i.e., put off or postponed health care, hesitant to get health care, did not come back for a follow-up appointment, did not follow the treatment plan or get a needed test, avoided the respective health care provider involved with the discrimination, and no longer used the respective health care facility where discrimination occurred).	Discrimination in health care was not significantly associated with not being current for mammography screening.
(Gonzales et al., 2014)	Discrimination based on being Native American/American Indian.	General health care settings.	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always).	A binary measure was created to categorize responses as ‘never’ (to all items) versus ‘any’ (rarely or more to at least one item).	Cronbach’s alpha = 0.94a	Discrimination in health care was associated with lower adjusted odds of dental exam; checks for blood pressure, creatinine, and total cholesterol; and pneumococcal vaccination. Any discrimination in health care was also associated with higher adjusted odds of having hemoglobin A1C values above target levels for diabetes control.	Discrimination in health care was not significantly associated with blood pressure (>130/80 mm Hg) or total cholesterol (>200 mg/dl).
(Greer et. al, 2014)	Discrimination based on being Blacka	Specific health care settinga	7 items rated on a 5-point scale (Always, Most of the Time, Sometimes, Rarely, Never)a	Items were reverse scored then summed, with higher scores indicating greater discrimination.	Cronbach’s alpha = 0.93	Discrimination in health care was significantly and positively correlated with health care mistrust (e.g., “people in my ethnic group should be suspicious of modern medicine”). Discrimination in health care was also significantly associated with decreased adherence to hypertension treatment (e.g., adherence to a low sodium diet). There was additionally a significant interaction between discrimination in health care and systemic racism on treatment adherence, such that treatment adherence was lowest for individuals with high discrimination in health care and low or average (versus high) systemic racism.	None reported.
(Hausmann et al., 2011)	Race- or color-based, and SES- or class-based discrimination	General health care settings	7 race-based items & 7 SES-based items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always)a	Responses were recoded as binary (i.e., ‘never’ versus ‘ever’) for each item, then summed for each item set, with higher scores indicating greater discrimination.	Cronbach’s alpha for race-based discrimination items = 0.93 Cronbach’s alpha for SES-based discrimination items = 0.90	Among African American participants, high levels of race-based discrimination in health care were significantly associated with less positive nonverbal affect tone on both the part of participants and their providers (during their health care interaction), as well as with low participant reports of provider warmth/ respectfulness and ease of communication. Among Whites, race-based discrimination in health care was negatively associated with verbal psychosocial exchange and with participant ratings of visit informativeness and ease of communication. Results were similar for SES-based discrimination.	Discrimination in health care based on race or color was not significantly associated with verbal measures of biomedical exchange, rapport building, or patient activation for African Americans and Whites. Among African Americans only, race or color based discrimination in health care was also not significantly associated with psychosocial exchange with providers. Informativeness of visit also was not significantly associated with race or color based or SES-based discrimination in health care for African Americans only. For Whites, provider warmth or respect were not significantly associated with race or color-based discrimination in health care. For both African Americans and Whites, SES-based discrimination in health care was not significantly associated with biomedical exchange, psychosocial exchange, rapport building, patient activation, positive patient affect, or positive provider affect. For Whites, discrimination in health care based on SES was not significantly associated with provider warmth or ease of communicating.
(Hausmann et al., 2010)	Race- or color-based discrimination	General health care settings	7 items on a 4-point scale (Never, Once, 2 or 3 Times, 4 Times or More)a	Responses were recoded as binary (i.e., ‘never’ versus ‘ever’) for each item, then summed. If participants reported ‘never’ to all measures, they were categorized as experiencing ‘none’ versus ‘any’ discrimination.	Cronbach’s alpha = 0.94	Discrimination in health care was associated with higher adjusted odds of experiencing problems with diabetes care.	Discrimination in health care was not associated with receiving all 5 recommended screenings for diabetes complications in the past 2 years.
(Jabson et al., 2011)	Sexual orientation-based discrimination	Health care settings where breast cancer treatments received	7 items rated on a 5-point scale (Response options were not reported in the article.)	Mean score across items were computed.	Cronbach’s alpha = 0.75	Discrimination in health care was significantly associated with quality of life.	None reported.
(Kressin et al., 2010)	Race-, color-, or ethnicity-based discriminationa	General health care settings	7 items rated on a binary scale (Yes/No)	The number of experiences reported was summed, with higher scores indicating more discrimination.	Cronbach’s alpha = 0.90	None reported. Discrimination in health care was not included in multivariable analysis of blood pressure control due to nonsignificant association found in bivariate analysis.	None reported. Discrimination in health care was not included in multivariable analysis of blood pressure control due to nonsignificant association found in bivariate analysis.
(Manze et al., 2010)	Race-, color-, or ethnicity-based discriminationa	General health care settings	7 items rated on a binary scale (Yes/No)	A binary measure was created to categorize anyone who answered “yes” to any item as having experienced discrimination.	Cronbach’s alpha = 0.90a	None reported.	Treatment intensification for hypertension was not significantly associated with discrimination in health care.
(Myaskovsk et al., 2011)	Race- or ethnicity- discriminationa	General health care settings	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always) a	Mean score across items was computed.	Cronbach’s alpha = 0.89	Discrimination in health care was associated with higher odds of occupational functioning (i.e., a person’s ability to participate in various activities such as employment or school).	Discrimination in health care was not significantly associated with physical independence, mobility, social integration, life satisfaction, general perceived health status, or current perceived health status.
(Myaskovskyet al., 2012)	Race- or ethnicity- discriminationa	General health care settings	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always) a	A binary measure was created to categorize responses as ‘any’ versus ‘none’ of the seven experiences.	Cronbach’s alpha = 0.90	Participants who reported any discrimination in health care took significantly longer to be accepted for a kidney transplant than those who did not report any discrimination in health care.	None reported.
(Myaskovsky et al., 2017)	Race- or ethnicity- discriminationa	General health care settings	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always)a	A binary measure was created to categorize responses as ‘ever’ versus ‘never.’	Cronbach’s alpha = 0.92	Overall, multivariable analysis results showed previous experience of discrimination was significantly associated with lower service satisfaction.	None reported. Discrimination in health care was not included in multivariable analyses of physical component summary of health, the mental component summary of health, and general health status due to nonsignificant associations found in bivariate analyses.
(Padela et al., 2014)	Religion-based discrimination	General health care settings	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always) a	Each response category was assigned a numerical value between 0 and 10. An average score was then computed if the respondent answered more than 50% of the items.	Cronbach’s alpha = 0.93	None reported.	Discrimination in health care was not significantly associated with cervical cancer screening status.
(Peek et al., 2011)	Discrimination based on race, ancestry, or national origin	General health care settings	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always)	The factor structure of measure items was explored by treating the measure items as continuous variables, and descriptive statistics were computed by taking the mean score across items.	Exploratory factor analysis results supported a single factor solution (eigenvalue = 4.36, accounting for 62% of the variance). All items had standardized factor loadings > 0.49. Cronbach’s alpha = 0.89 (original sample) & 0.85 (retest sample) Test-retest reliability = 0.58, p < .001 Correlations with other variables indicated measure’s convergent validity and discriminant validity.	None reported.	None reported.
(Sheppard et al., 2017)	Race- or ethnicity-based discrimination a	General health care settings	7 items rated on a binary scale (Yes/No)a	A binary measure was created to categorize responses as ‘any’ versus ‘none’ of the experiences.	Cronbach’s alpha for entire sample = 0.87, for Black women = 0.86, and for White women = 0.81a	None reported.	Discrimination in health care was not significantly associated with chemotherapy use.
(Sheppard et al., 2011)	Race- or ethnicity-based discriminationa	General health care settings a	7 items rated on a binary scale (Yes/No)a	Mean score across items was computed.	Cronbach’s alpha = 0.81	None reported.	None reported.
(Sheppard, O’Neill, et al., 2015)	Race- or ethnicity-based discrimination a	General health care settings	7 items rated on a binary scale (Yes/No)a	A binary measure was created to categorize responses as ‘any’ versus ‘none’ of the experiences.	Cronbach’s alpha for entire sample = 0.87, for Black women = 0.86, and for White women = 0.81a	None reported. Discrimination in health care was not included in multivariable analyses of RS testing and chemotherapy use due to nonsignificant associations found in bivariate analysis.	None reported. Discrimination in health care was not included in multivariable analyses of RS testing and chemotherapy use of blood pressure control due to nonsignificant association found in bivariate analysis.
(Sheppard, Oppong, et al., 2015)	Race- or ethnicity-based discrimination a	General health care settings	7 items rated on a binary scale (Yes/No)a	A binary measure was created to categorize responses as ‘any’ versus ‘none’ of the experiences.	Cronbach’s alpha for entire sample = 0.87, for Black women = 0.86, and for White women = 0.81a	None reported. Discrimination in health care was not included in multivariable analyses of time to surgery due to nonsignificant association found in bivariate analysis.	None reported. Discrimination in health care was not included in multivariable analyses of time to surgery due to nonsignificant association found in bivariate analysis.
(Sheppard et al., 2008)	Discrimination based on Latino/Hispanic ethnicity	General health care settings	6 items rated on a binary scale (Yes/No)	A binary measure was created to categorize responses as ‘any’ versus ‘none’ of the experiences.	Cronbach’s alpha = 0.60	Participants who had not experienced discrimination in health were more likely to report satisfaction with health care relationships.	None reported. Discrimination in health care was not included in multivariable analyses of recent mammography due to nonsignificant association found in bivariate analysis.
(Sheppard et al., 2014)	Ethnicity-based discrimination	General health care settings	6 items rated on a binary scale (Yes/No)	A binary measure was created to categorize responses as ‘any’ versus ‘none’ of the experiences.	Not reported	Participants whose home language was Spanish and English (versus Spanish only) had higher adjusted odds of discrimination in health care. Additionally, higher health care satisfaction and better communication with providers were each associated with lower adjusted odds of discrimination in health care.	Discrimination in health care was not significantly associated with education level, health insurance status, provider ethnicity (i.e., Latino or non-Latino), or trust in provider.
(Vu et al., 2016)	Religion-based discrimination	General health care settings	7 items rated on a 5-point scale (Never, Rarely, Sometimes, Most of the Time, Always) a	Each response category was assigned a numerical value between 0 and 10. An average score was then computed if the respondent answered more than 50% of the items.	Cronbach’s alpha = 0.93	None reported.	Discrimination in health care was not significantly associated with delayed care seeking due to a perceived lack of female clinicians.

^aInformation was obtained directly from authors.