Family Physician Racial Identity: An Analysis of “Other” Race Selection and Implications for Future Data Collection

• Cultural Diversity
• Data Collection
• Family Medicine
• Family Physicians
• Race
• Workforce

While race is a social, cultural, political, and economic construct, collecting data that attempts to categorize people along racial lines is necessary to uncover the very real effects of racism, including racialized health disparities.¹ The categories used to capture race on the United States Census have changed over time according to geopolitical events and shifting beliefs and attitudes and starting in 2000, respondents were allowed to select more than 1.² The American Board of Family Medicine (ABFM) began collecting race and ethnicity of its Diplomates in 2013, for the primary purpose of conducting Differential Item Functioning (DIF) analyses of certification examination items to ensure that no examination items are inherently biased.³ We sought to understand the hidden diversity of physicians who selected “Other” race to inform data collection improvements.

Between 2016 (when the “Other” category was added) and 2021, 64,067 family physicians have responded to the question, “Select the race with which you most identify (select 1)”, which, in addition to the 5 basic categories (American Indian or Alaska Native; Asian; Black or African American; Native Hawaiian or Other Pacific Islander; White), includes an “Other” option with an open-text box. Hispanic ethnicity is captured in a separate item with just 2 response options. Using a qualitative content analytic approach,⁴ we systematically manually categorized all open-text responses from the “Other” race selection into mutually exclusive categories, using the US Census Bureau’s 2020 classification standards³ and ongoing research on how to improve data collection.⁴ We then calculated simple frequencies and proportions for each category.

A total of 4519 Diplomates (7.1%) chose “Other.” As shown in Table 1, 10 categories emerged. Four racial/ethnic categories (Hispanic, South Asian, Multi-racial, and Middle Eastern/North African [MENA]) make up the majority (66.8%) of these comments. Apart from the 9.7% who wrote cynical, critical, or unclear remarks and the 9.6% who did not provide a response at all, the remaining “Other” comments included national or regional identification (8.6%), preference not to respond or no racial category applies (4.5%), and religious–ethnic identification (0.9%).

Our analysis of the open text responses provided by Diplomates who select “Other” as a racial category they most identify with, aligns with findings from the census² and suggests that the ABFM should revise the way it collects race and ethnicity data. First, similar to previous studies, the large number of “Hispanic” entries demonstrates many Diplomates do not identify as a single racial category separate from their Hispanic identity; a single item that includes “Hispanic” as an option should be used.^2,5 Second, rather than selecting “Asian,” many choose “Other” and write in a more detailed answer; providing more granular Asian categories will help ensure that the diversity of “Asian” physicians is recognized.⁵ Similarly, the addition of MENA as an option is new best practice and would make visible this growing group of physicians.^2,5,6 Finally, for the Diplomates who identified as more than 1 racial category, allowing ‘Select all that apply’ is the method consistently utilized as best practice, and will capture more accurate and meaningful data.⁵ These proposed changes will capture higher quality, more accurate data that better reflect racial identities. This will improve the ability to detect potential bias in examination items and create a more nuanced understanding of the family physician workforce, which can inform more effective policies and programs toward diversity, equity, and inclusion.

• Received for publication January 24, 2022.
• Revision received March 3, 2022.
• Accepted for publication March 8, 2022.