The US population is becoming increasingly diverse. Racial and ethnic health disparities and inequities have been extensively documented. As a discipline, family medicine can help mitigate these disparities by fostering a physician workforce that is reflective of the diverse patient population it serves, since patient-physician racial concordance leads to better communication, medical adherence, and patient satisfaction.¹

Currently, the American Board of Family Medicine (ABFM) collects race and ethnicity of its diplomates (a proxy of the family medicine physician workforce) through 5 broad racial categories (American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White) and 2 ethnicity categories (Hispanic or non-Hispanic). Outside of these set categories, there is an “Other” option that allows open-text responses.²

These large 5 racial categories stem from the revised 1997 Office of Management and Budget (OMB) standards for the classification of federal data on race and ethnicity.³ The standards were meant to serve as a minimum set of racial and ethnic categories but do not prohibit the collection of additional detailed categories. In fact, the OMB standards encourage greater detailed data when useful and so long as the additional details can be aggregated back to the minimum set standard categories. Unfortunately, these categories are often misinterpreted as the only permissible reporting categories, serving as an unnecessary barrier to fully understanding the makeup of the family medicine workforce.

In the article “Family Physician Racial Identity: An Analysis of ‘Other’ Race Selection and Implications for Future Data Collection,” Eden et al.² analyzed these “Other” open-text responses from 2016 to 2021. That timespan yielded 64,067 family medicine diplomates with a total of 4519 diplomates (7.1%) who selected “Other.” Results through a qualitative thematic analysis of these responses showed that most of them were of racial/ethnic categories that the respondents felt were not adequately represented in the given 5 racial categories.² Accordingly, this suggests that the ABFM could advance diversity, equity, and inclusion through meaningful disaggregation of its racial and ethnic data collection. Data disaggregation is important because data gets broken down into detailed subcategories to reveal inequities that may otherwise be hidden by the current reporting systems.

Significant heterogeneity exists within each of these broad racial/ethnic categories. There lies a diverse array of nationalities, languages, immigration and refugee histories, nativity status, socioeconomic backgrounds, and levels of acculturation. When race/ethnicity data are collected only within large categories, a sufficient level of detail is lacking to make nuanced distinctions regarding the health and social needs of different racial and ethnic groups. Indeed, data inequity, or the continued invisibility and lack of representation in data for certain populations, particularly those with significant disparities, is a critical source of health inequity.⁴

For example, how might the factors that determine health outcomes differ for a tenth generation English American compared with those for a Syrian refugee? Recent immigrants from the Middle East and North Africa (MENA) needing to migrate due to civil unrest face social, economic, and political climates that get overlooked when they are broadly categorized as “White.”⁵ Likewise, the “Asian” category encompasses all who originate from the Far East, Southeast Asia, or the Indian subcontinent³ but do little to acknowledge whether migration patterns stem from a pursuit of economic opportunity versus from a necessity to resettle as refugees from plights of war. Grouping all African-descent or Afro-Caribbean populations into the “Black or African American” category without regard to country-of-origin or nativity also masks differences in documented health and socioeconomic characteristics.^6⇓–8

In response to the call for more detailed, disaggregated data over the years, the United States Census Bureau has implemented more subcategories into the US Census. The ABFM could consider the standards used in the 2020 decennial census as a useful reference. For instance, the “Asian” category in the US Census includes subgroups like “Filipino” and “Vietnamese” and within the “Pacific Islander” category, subgroups like “Samoan” and “Chamorro.” Individuals who identify with more than 1 race are also given the option to choose multiple races in response to the race question.⁹

However, while the US Census may serve as a potential starting point, there are at least 2 instances in which current federal government methods for race and ethnicity data collection are inconsistent with certain groups’ lived identities. To begin with, the terms “Hispanic” and “Latino” are uniquely American creations that have no real meaning in Latin America. The social construct of race in the United States rests on a person’s physical appearance, differing greatly from the Latin American understanding which involves concepts like birthplace and culture. Latinos familiar with the latter understanding of race may not then identity with the US racial system and may often select “Other” as their race. Furthermore, although current federal standards classify “Hispanic or Latino” as an ethnic identity, Latinos are often treated as a racial group in US society.¹⁰

Similarly, while the federal government considers those with Middle East and North Africa (MENA) origins as “White,” a recent study suggests evidence that a majority of MENA individuals do not see themselves as “White,” nor do a large percentage of people who identify as “White” without MENA origins perceive MENA people as “White” either.¹¹ While there is a possibility for a MENA option in the 2030 US Census after thousands of supportive public comments to the OMB and the Federal Interagency Working Group for Research on Race and Ethnicity,^12,13 results from Eden et al.² reveal that ABFM diplomates call for a separate MENA identity category now. Otherwise, many potential disparities and inequities faced by MENA Americans will continue to remain hidden.

Certainly, developing an exhaustive list of racial and ethnic categories can be overwhelming and impractical. However, it is vital to determine the underlying purpose and usage of the racial and ethnic data and to have that guide the data collection. As disaggregated data are being collected for the general population in the US Census, the field of family medicine ought to know if its workforce accurately resembles the patient population and if not, to figure out which communities are missing representation. How many of our physicians come from underrepresented communities that have not been traditionally seen as underrepresented due to aggregated data? How many of these invisibly underrepresented colleagues might have been unintentionally denied allocation of much needed resources and mentorship? How can we ensure our diverse patient population receives care from an equally diverse physician workforce?

Successfully eliminating racial and ethnic disparities and inequities will be extremely difficult to achieve without a meaningful system of data collection and analysis with which to measure and track progress. Certainly, delaying action on the appropriate collection and use of racial and ethnic data will only guarantee that opportunities to improve diversity, equity, and inclusion in the family medicine workforce will continue to be missed. Lest many of those underrepresented in our community will remain invisible, future data collection must be done with a careful and deliberate focus on equity.