Aspiring to equity: race vs ethnicity in clinical research

In this issue of Blood Advances, Nze et al¹ explore the impact of social determinants of health on clinical trial participation in non-Hodgkin lymphoma and provide an example of a center deliberately investing in strategies to improve diversity in clinical trial participation. The results suggest that the racial composition of clinical trials was not associated with differences in trial participation. This raises the possibility that systemic racism in trial recruitment and, therefore, clinical research more broadly, can be mitigated. This is critical to aspire because once data are generated with blindness toward a specific population, this bias cannot be easily undone and predicates future research.

According to the World Health Organization, social determinants of health can be more important than health care or lifestyle choices, accounting for 30% to 55% of health outcomes.² These are the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems that shape daily life. These include economic policies and systems, development agendas, social norms, social policies, and political systems.² Population diversity in clinical trials has historically been challenging. Ideally, clinical research advances the knowledge of the biology of disease and the efficacy and toxicity of therapies to inform treatment decisions. However, if affected populations with potential biological differences such as age, sex, and ethnicity are not represented, they are deprived of equity in the quality of the evidence informing their care. Efforts to mitigate health care disparities through health services and clinical research are critical for improving health care access and equity of outcomes.

This is supported by the study by Nze et al, which demonstrated that many social determinants of health are challenging to capture in clinical practice and research. This results in the pragmatic use of approximate surrogate metrics, such as address, health insurance status, and composite estimates, such as neighborhood socioeconomic status and the area deprivation index.¹ Although sociocultural constructs are difficult to quantify and are often not prospectively collected in clinical trials, biological constructs that intersect with disparities are within reach. It is critical to be able to distinguish these in clinical trial design. There is increasing awareness of sex-based biological variables,³ and that self-identified gender may differ from biological sex. This is an important distinction that allows the collection of sex-based variables in clinical trials with the inclusion of data on patients whose gender is not their biological sex who are marginalized, but also that facilitates the precision of data collection as well as its future applicability. Similarly, there is a difference between race and ethnicity.⁴ These are often conflated or used interchangeably, leading to misconceptions and potentially biased conclusions and applications of data.

As noted by Nze et al, race refers to a socially constructed classification system based on physical characteristics, such as skin color, facial features, hair texture, or stature, and self-identification is often used as the primary method for capture. Data on race in clinical research are therefore a reflection of power dynamics, social structures, and historical contexts that result in inequities in health care.¹ The definition of race also differs by country or region. It is important to recognize that race is a source of health care disparity and that this data point is a measure of racism⁵ and not specific biological differences because race is not a biological construct.

In contrast, ethnicity encompasses shared cultural, linguistic, or ancestral characteristics among a group of people. In clinical research, the implication is potential biological differences in pharmacogenomics. This is important because it is estimated that 20% to 30% of the variability in drug response can be explained by genetic polymorphisms in genes involved in drug absorption, distribution, metabolism, and excretion, as well as drug target and immune-related genes.⁶ The impact on therapeutic efficacy and toxicity occurs in up to 50% of treatments, and adverse drug reactions account for 6% to 9% of all hospital admissions, of which up to 40% are life-threatening.⁷ These adverse outcomes can also be further exacerbated by social determinants of health. Integrating pharmacogenomics into clinical research is a more accurate and effective way to manage these risks.

Nze et al examined local practices that identified areas of potential improvement. For instance, Hispanic patients had lower odds of participation than White patients in their diffuse large B-cell lymphoma cohort, whereas overall, race, sex, insurance type, and neighborhood socioeconomic status were not associated with trial participation.¹ The need to address this specific population disparity would have been more difficult to detect without intentional examination. This study exemplifies a quality improvement approach to equity in clinical trial recruitment that facilitates the allocation of resources to most efficiently close disparity gaps.

Clinical research design needs to prospectively and thoughtfully collect data on social determinants of health with an intersectional lens⁸ and with as much specificity as possible from the outset. This would allow the analysis of factors that may have a confounding effect. Data on ethnicity, gender, and other intersectional characteristics should be collected uniformly; organ function variables that are inclusive of ethnic variances should be used; and pharmacogenomics studying efficacy and toxicity need to become the new normal. For example, it would be of clinical relevance and value to understand the outcomes of a Black woman to appreciate the intersection of racism and gender for the therapy in question, or an Asian woman with a specific pharmacogenomic profile, to better understand the intersection of race, pharmacology, and gender. Finally, wherever possible, it is necessary to acknowledge the limitations of the data that are not generated from diverse populations as they apply to ethnicity and ensure that previous assumptions are re-examined and answered in future research. An intersectional approach⁸ is likely to be the most efficient way to achieve this, but it begins with a deliberate intention to mitigate bias, the implementation of strategies to deliver equity, diversity, and inclusion,⁹ and a continuous improvement mindset. Ultimately, integrating equity, diversity, and inclusion in all aspects of clinical research is central to high-quality clinical research.

Conflict-of-interest disclosure: N.H. declares no competing financial interests.