3

Moving from Race-Based to Race-Conscious Medicine to Address Health Inequities

Justin List, director of health care outcomes in the Veterans Health Administration Office of Health Equity, defined race-based medicine as a system that treats race as a biological variable and, when translated into clinical practice, leads to racial health inequities. “Race-based medicine fails to recognize race as a sociopolitical construct,” List said. He also defined race-conscious medicine as medical practice and pedagogy that accounts for how structural racism determines illness and health outcomes (Cerdeña et al., 2020). List noted that treating race as biological essentialism, and conflating race, genetics, ancestry, and medical language, has harmed minoritized patients.

RACE AND eGFR: LESSONS LEARNED AND FUTURE DIRECTIONS

Amaka Eneanya, adjunct associate professor of medicine at the University of Pennsylvania, said the U.S. Renal Data System annual data report for 2022 showed that Black individuals have a higher prevalence of chronic kidney disease than White individuals and other racial and ethnic groups. The data also show increasing prevalence in chronic kidney disease since 2010 and indicate that more advanced stages of chronic kidney disease disproportionately affect Black individuals (U.S. Renal Data System, 2022).

One reason for this inequity, said Eneanya, is that diabetes and high blood pressure—the top two causes of chronic kidney disease in the

United States—are more prevalent in Black individuals than in White individuals. Black individuals are also less likely to receive nephrology care before starting dialysis than other racial and ethnic groups. Moreover, the risk of developing end-stage renal disease is four times higher in Black individuals and 1.3 times higher in Hispanic individuals than in White individuals (U.S. Renal Data System, 2022). At the same time, Eneanya noted, Black individuals are less likely to receive a kidney transplant than any other racial and ethnic group in the United States.

Eneanya said social determinants of health are also responsible for inequities related to chronic kidney disease. Structural racism, lifestyle issues (e.g., diet and housing insecurity), and racial discrimination can influence allostatic load and other biological manifestations of stress, which in turn affect kidney health. Other studies found allostatic load is higher among Black individuals than among White individuals (Duru et al., 2012; Rogers et al., 2022), which can lead to neurochemical and metabolic changes that reduce the estimated glomerular filtration rate (eGFR).

The eGFR equation adjustment that Vyas discussed results in clinicians categorizing Black individuals as having better kidney function than is accurate. Eneanya noted this adjustment arose from a 1999 finding that Black men and women had higher levels of serum creatinine for any measured level GFR than all other racial and ethnic groups (Levey et al., 1999). To account for that difference, investigators introduced an adjustment based on the debunked fallacy that Black individuals have naturally higher serum creatinine levels because they have more muscle mass, independent of physical activity. Because of this racial adjustment, Black individuals experience delays in receiving critical preventive care and increased transplant waitlist time, Eneanya said.

Race-based eGFR adjustments exacerbate disparities by characterizing Black individuals as having better kidney function than other racial groups. For instance, Eneanya pointed out that Black individuals are more likely than White individuals to receive dialysis care via a catheter rather than a fistula; however, the latter produces better health outcomes and less morbidity and mortality. Although national and international guidelines recommend early fistula placement for patients with advanced kidney disease, race-based eGFR equations may delay this procedure among Black patients (Eneanya et al., 2022). Importantly, race-based eGFR lacks an accommodation for mixed raced individuals. Eneanya recalled how one biracial individual, who identified as a Black woman and had one Black parent, asked her nephrologist to list her as White, so her eGFR score would qualify her for the transplant waitlist. Another issue Eneanya raised was transparency during shared decision making—particularly when clinicians fail to tell Black patients they are using race to make clinical decisions. Eneanya blamed explicit bias for the use of race to judge a patient’s clinical pathway.

In September 2021, the National Kidney Foundation and American Society of Nephrology Joint Task Force issued a recommendation to immediately implement a newly developed race-neutral equation for determining eGFR levels (Delgado et al., 2021, 2022; Inker et al., 2021). The task force also called for national efforts to increase the use of other biomarkers, such as cystatin C, to assess kidney function and to confirm eGFR readings in adults who are at risk for or who have chronic kidney disease (Delgado et al., 2021, 2022; Inker et al., 2021). Then, in June 2022, the Organ Procurement and Transplantation Network issued a policy prohibiting all U.S. transplant centers from using any race-based eGFR equation for kidney transplant listing. A second policy, issued in January 2023, required all U.S. kidney transplant centers to review all Black patients on the kidney transplant waitlist and to modify their waitlist time if their original listing test used a race-based algorithm.¹ Eneanya cautioned that changing one equation will not eliminate all racial disparities in treating chronic kidney disease and that other actions are needed.

She also highlighted barriers to implementing the new race-neutral algorithm. A survey conducted in September 2021 by the Chronic Kidney Disease Epidemiology Collaboration found that only 31 percent of U.S. laboratories had switched to the race-neutral eGFR algorithm, and 32 percent reported they did not plan to adopt the new algorithm (Genzen et al., 2022). Eneanya noted continued resistance among much of the medical community regarding race-free algorithm accuracy. In addition, “There has been bullying and gaslighting toward individuals, particularly young individuals, who are pushing for change. I get notifications on a monthly basis that people are using their power to slow or prevent this change,” Eneanya said.

IMPLEMENTATION OF RACE-FREE EQUATIONS

Paul Palevsky, deputy executive director of Veterans Health Administration Kidney Medicine Program, said it often takes 10 years or longer to implement a new recommendation. Regarding the eGFR equation, National Kidney Foundation began working with commercial and health system laboratories and pathology societies to prepare them for change prior to releasing its recommendation to use the new race-free equation.

According to Palevsky, one barrier to change has been Logical Observation Identifiers Names and Codes (LOINC codes). Any change in laboratory data requires generating a new LOINC code designation for

___________________

¹ https://optn.transplant.hrsa.gov/news/optn-board-approves-waiting-time-adjustment-for-kidney-transplant-candidates-affected-by-race-based-calculation (accessed September 29, 2023).

laboratory results produced using the new equation and inserting it into the electronic health record (EHR), he said. As a result of work prior to publishing the new race-free equation, Regenstrief Institute, which generates LOINC codes, issued prerelease special use codes on October 1, 2022, as a first step toward implementation; this occurred about 1 week after the equation’s release. Palevsky and his colleagues developed guidance materials on the new equation and codes for clinicians, laboratories, and patients. Meanwhile, the American Association of Clinical Chemistry and the National Kidney Foundation published a guidance document on improving equity in chronic kidney disease care and implementing race-neutral equations (Pierre et al., 2023).

Palevsky said the survey Eneanya mentioned found that much work remains before full implementation in all U.S. laboratories is achieved. However, when the survey was conducted—less than 6 months after the equation’s release—more than three-quarters of the laboratories already knew about the new equation (Genzen et al., 2022). He noted that four laboratories, which account for more than 50 percent of the 250 million creatinine assays and associated eGFR levels reported annually, had deployed the new equation by the end of summer 2022. In addition, the Veterans Health Administration issued a directive in January 2022 that its laboratories needed to implement the new equation by April 1, 2022 (Crowley, 2022).

Palevsky said removing race from eGFR estimations required a carefully organized implementation strategy that included early engagement of the laboratory medicine community, rapid release of new LOINC codes, multidisciplinary clinical education, and patient education.

RACE IN PULMONARY FUNCTION TEST INTERPRETATION: HISTORY, CLINICAL IMPLICATIONS, CHALLENGES, AND ALTERNATIVES

Nirav Bhakta, assistant professor at University of California, San Francisco, said the pulmonary function community has lagged behind other areas of clinical medicine in acknowledging that its assessment of patient lung function was racist and relied on racialized tests. He explained that a racially centric organization of pulmonary function data results in a residual difference in lung function between White people and people from Southeast Asian or African American populations. This differential implies that an African American individual with a given forced expiratory volume (FEV)—a measure of how much air a person can exhale during a forced breath—would be rated with a higher lung function than a White individual with the same FEV value. Thus, using a race-based equation directly affects a clinician’s assessment of whether an

individual has a respiratory disease, whether additional tests are needed, and whether the patient should be referred for a lung transplantation evaluation. Using the race-based equation can also impact determinations of physical eligibility to work as a firefighter or a diver and impact disability determinations.

Bhakta then discussed examples of challenges in using race to make clinical decisions: Race has not been defined uniformly across time and location, and race can conceal modifiable risks for reduced lung function and disease. In addition, racial categories are limited, and existing ones overlook the effects of acculturation and mixed backgrounds. The current practice for interpreting lung function measurements, Bhakta explained, involves adjusting results for age, height, sex, and unlimited categories of race and ethnicity but disregards the effect of social determinants of health. The measurement is then used to triage respiratory symptoms, diagnose conditions, determine disability, decide prognosis, evaluate the risk/benefit ratio for various treatments, and develop inclusion criteria for clinical trials.

A 2023 literature review identified many gaps and assumptions that informed the use of race and ethnicity in pulmonary function test interpretation (Marciniuk et al., 2023). A project Bhakta and his colleagues conducted produced an official statement for the American Thoracic Society that recommended no longer using race- or ethnicity-based pulmonary lung function test interpretation. Instead, they recommended using an average reference equation for all people (Bhakta et al., 2023). This project also identified many determinants of pulmonary function, including prematurity, infection, nutrition, exposure to second-hand smoke and other indoor and outdoor air pollution, discrimination, stress, and genetic factors. The recommendations relied on analyzing data from a variety of sources that show race-neutral equations more accurately reflect respiratory symptoms and mortality than race-based equations (Baugh et al., 2022; Elmaleh-Sachs et al., 2022; McCormack et al., 2022). Bhakta said the studies underpinning the new recommendations show that using a single reference equation is superior to using race-based equations when the goal is using pulmonary function test measurements to equitably apportion clinicians’ attention and medical resources to people with similar risk of bad outcomes.

Bhakta then indicated several factors that affect implementation challenges. For example, updating lung function data acquisition and interpretation software is expensive. In addition, “we need to increase our ability to speak with patients and colleagues about these changes and address their concerns about the impact of changes on White people, employment for Black people, and for people that are concerned about precision, again, highlighting the illusion of precision that racial ethnic

categories give rather than a true precision,” Bhakta said. He further highlighted the importance of continued engagement with the many people outside of clinical medicine who use pulmonary function testing and rely on its measurements for disability and insurance evaluation, occupational evaluation, and life insurance eligibility.

According to Bhakta, many challenges remain. For example, the ideal reference population is uncertain; approaches not requiring reference equations are needed; the value of FEV₁ is decreasing as new lung function tests are developed; optimal values for measurements related to body size, task, or exposure are difficult to determine; and uncertainties about how to adjust for chest dimension or environmental influences exist. In addition, the current thresholds for decisions based on pulmonary function tests largely lack empirical support. A need also exists to anchor the use of such tests on patient-centered outcomes rather than on a normal–versus–abnormal designation when comparing to reference values.

RACE-BASED OR RACE-CONSCIOUS: EXPLORING THE UTILITY AND FUNCTION OF SCREENING FOR BODY MASS INDEX IN ASIAN AMERICAN COMMUNITIES

Nadia Islam, associate director for New York University (NYU) Langone Health’s Institute for Excellence in Health Equity, discussed the utility and function of body mass index (BMI) as a clinical screening tool for Asian American communities. In June 2023, the American Medical Association (AMA) released a policy statement noting several problems with the clinical use of BMI, including that BMI fails to differentiate between body fat and lean body mass, nor does it account for life cycle, body fat location, or the location of accumulated fat caused by hormones (AMA CSAPH, 2023). The policy statement notes problems with BMI as a predictor of morbidity and mortality. For example, correlating mortality rates with BMI fails to account for factors such as family history or genetics. Perhaps most important, said Islam, AMA highlights the eugenics history behind BMI and how it has been used for racist exclusion.

Islam argued that the relationship between BMI, stigma, and discrimination has manifested differently for Asian American populations as a reflection of two distinct but mutually reinforcing myths: that Asian Americans are both the model minority and the perpetual foreigner. “This context is important,” said Islam, “because it is both driven by and results in the lack of data equity in Asian American communities. Data aggregation within and across Asian American populations, which has lumped together more than 30 unique subgroups, renders our communities invisible.”

As Islam and her colleagues have stated, “there is a mutually reinforcing cycle of poor data quality and racialized stereotypes that shapes our

understanding of Asian American health” (Yi et al., 2022). This community invisibility, said Islam, has occurred even though Asian American populations represent the fastest growing minoritized group in the United States.² While “Asian American” is an umbrella term, it encompasses many ethnicities with different migration experiences, socioeconomic statuses, and language proficiencies. In addition, said Islam, Asian American populations are and continue to be the most understudied U.S. racial or ethnic group. Despite inadequacies in data collection and reporting, an analysis of National Health and Nutrition Examination Survey data from 2011 to 2016 found that Asian Americans as a group have higher rates of both diagnosed and undiagnosed diabetes than all other racial and ethnic groups. Disaggregated data also show stark differences across Asian American population subgroups, including high rates among particular groups (e.g., those from South and Southeast Asia) (Cheng et al., 2019).

Though rates of undiagnosed diabetes are decreasing nationally, Asian American populations continue to have the highest rates of undiagnosed diabetes (Fang et al., 2022). Islam said diabetes risk and BMI have a unique relationship in Asian American populations. Asian American individuals develop diabetes at lower BMI values than individuals from other racial and ethnic groups. One analysis found that screening for diabetes at a BMI of 25—values above 25 are considered overweight—left a large proportion of Asian Americans in the undiagnosed category (Hsu et al., 2015). Because of this miscategorization, a perception exists that Asian American populations experience a lower risk for developing diabetes and therefore need to be screened less often, even though Asian American populations have higher rates of diabetes. Responding to this finding, the American Diabetes Association (ADA) issued a position statement recommending that treatment guidelines include screening for diabetes in Asian Americans with a BMI of 23 or higher (ADA, 2015). In addition, the National Council of Asian Pacific Islander Physicians launched the Screen at 23 initiative to promote diabetes screening among Asian American populations.³ The U.S. Preventive Services Task Force adopted this lower BMI cutoff for Asian American populations in 2021 (USPSTF, 2021).

Islam and her colleagues operationalized this recommendation in New York City by implementing a multilevel initiative in primary care practices serving a large South Asian population. The initiative had two

___________________

² https://www.pewresearch.org/short-reads/2021/04/09/asian-americans-are-the-fastest-growing-racial-or-ethnic-group-in-the-u-s/#:~:text=Asian%20Americans%20are%20the%20fastest,ethnic%20group%20in%20the%20U.S.&text=Asian%20Americans%20recorded%20the%20fastest,States%20between%202000%20and%202019 (accessed September 29, 2023).

³ Additional information is available at http://ncapip.org/diabetes/screenat23/index.html (accessed September 29, 2023).

key components: working with practices to enhance their capacity to use the EHR to better screen and identify patients with prediabetes and linking patients to community health workers who provide culturally tailored health coaching and education to support lifestyle changes. Preliminary results indicate that individuals who received the intervention were more likely to lose more weight than those who were in the control group. These results, said Islam, underscore that models pairing screening at BMI 23 with culturally tailored coaching can effectively promote diabetes prevention in South Asian communities.

In contrast to the potential utility of Screen at 23, a seminal paper in the field (Cerdeña et al., 2020) indicated Screen at 23 as an example of potentially harmful race-based medicine. Islam argued, “We must consider that enhanced screening is a different proposition than imposing differential diagnostic race-based criteria.” She continued, “Although the paper cites the potential for stigma associated with increased screening, I think this phenomenon needs to be better investigated with input from communities.” Islam’s research, which has used community-engaged and partnered approaches, has demonstrated Asian American communities are receptive to increased screening and want diabetes resources that are culturally and linguistically tailored. Providing evidence-based interventions and self-management resources relevant to communities can mitigate any potential stigma associated with increased screening.

One recent analysis found that screening for diabetes and prediabetes at lower age and BMI thresholds resulted in greater sensitivity and lower specificity, particularly among Hispanic, non-Hispanic Black, and Asian adults (O’Brien et al., 2023). The analysis also demonstrated that screening all adults ages 35 to 70, regardless of BMI, resulted in the most equitable performance across all racial and ethnic groups. This suggests a need, argued Islam, for implementing more universal diabetes screening starting at younger ages across racial and ethnic groups.

Islam said there is a need to consider an evolution of the application of BMI as screening criteria and to understand the historical context of ADA recommendations and Screen at 23 campaign guidelines. Screen at 23 was initiated because data demonstrated the possibility of better optimizing BMI cutoff points when screening Asian American populations for diabetes and because the historical practice of using a BMI cutoff of 25 left a large proportion of Asian Americans undiagnosed. The key, said Islam, is that screening cannot occur in a silo and must be paired with culturally tailored resources and evidence-based interventions. She noted two additional needs: increasing physician education on the variation of risk differences among ethnic and age groups at varying BMI scores and using BMI with other valid measures of risk, such as visceral fat, body adiposity, and waist circumference.