Designing Evidence That Reflects Everyone: A Black History Month Perspective

Diversity in clinical trials is not only a social imperative, it is a scientific standard. February, during Black History Month, offers an important moment to reflect on both the historical context and the ongoing evidence gaps that persist in clinical development.

Black Americans have been historically underrepresented in clinical studies informing healthcare policy and drug development. Despite making up 14% of the U.S. population, Black individuals represent just 5% of clinical trial participants[1]. Considering the lower life expectancies in black Americans vs white Americans (72.8 vs 77.5)[2], steps must be taken to improve health equity in clinical research.

This represents a failure of both social equality and scientific rigor, because evidence that does not reflect the intended patient population cannot be considered fully reliable. When trial populations do not accurately reflect the target population’s demographics, it reduces clinical trial generalizability, undermines regulatory confidence, and limits healthcare applications[3].

Diversity in Clinical Trials as a Scientific Standard

Black participants have been historically mistreated in clinical trials, such as the notorious U.S. Public Health Service Syphilis Study at Tuskegee from 1932 to 1972. In the study, black men were misled about receiving treatment, and effective therapy was withheld even after it became available. The resulting harm remains one of the most cited examples of ethical failure in U.S. medical research.

Tuskegee is an infamous case that reflects a broader pattern of exclusion and mistrust that has affected clinical research for decades[4]. The consequences are profound; significantly fewer Black individuals participate in clinical trials.

As a result, in 2021, the U.S. Preventive Services Task Force indicated it was unable to develop specific colorectal cancer screening guidelines for Black individuals, despite them having the highest incidence and mortality rates of the disease[5]. 

The scientific consequence is equally clear: when trial populations do not reflect real-world demographics, the validity and generalizability of study findings are compromised. Clinicians must ensure that studies accurately reflect ethnic variations in drug metabolism, disease prevalence, and treatment response to improve healthcare equity.

Regulatory bodies have recognized the importance. The Food and Drug Omnibus Reform Act of 2022 requires sponsors of Phase III and pivotal studies to submit Diversity Action Plans to the U.S. Food and Drug Administration (FDA) that include enrollment goals disaggregated by age, sex, race, and ethnicity[6].

Enrollment is only the Beginning: Performance Must Be Validated Across Populations

Ensuring representative enrollment is a vital first step. However, enrollment alone does not guarantee that data are equally reliable across participants. Measurement tools, sensors, and analytic frameworks must demonstrate consistent performance across diverse populations to preserve scientific integrity.

Device Validation Across Populations

The clearest example is the pulse oximeter, which estimates blood oxygen saturation by illuminating the skin. Because melanin absorbs light differently, pulse oximeters can systematically overestimate oxygen saturation in individuals with darker skin tones, introducing measurement bias that can mask clinically significant hypoxemia[7].

During the COVID-19 pandemic, this had direct consequences on Black healthcare, where pulse oximeter inaccuracies in dark-skinned patients were associated with delayed COVID-19 therapy, as patients who appeared stable on the monitor were in fact deteriorating[8]. In response, the FDA convened an advisory panel to address bias in digital technology, which guided manufacturers to validate devices across skin tones.

Enrollment diversity without analytic validation introduces the risk of systematic measurement error within subpopulations. When digital biomarkers are used as endpoints, such bias can materially affect trial conclusions.

The Monk Skin Tone scale has been recommended and is now largely used; it offers greater granularity (10 shades) than the traditional Fitzpatrick scale (6 shades), particularly for darker skin tones[9]. Ensuring digital health technologies are fit for purpose requires analytic validation through frameworks such as V3+, supported by ongoing research into sensor performance and algorithm accuracy.

The challenge is not limited to pulse oximetry. A review of wrist-worn wearable devices suggested that heart rate measurement accuracy is lower in darker-skinned individuals than in those with lighter skin tones[10]. As wearable-derived digital biomarkers increasingly serve as primary and secondary endpoints in clinical studies, ensuring consistent performance across skin tones is vital.

Algorithmic Bias and Digital Health

Machine-learning algorithms are trained on datasets that define their performance boundaries. If those datasets are misrepresentative of the population's diversity, the algorithm itself can become biased. Underrepresentation of darker-skinned patients in training datasets has led to a lower performance of AI-based skin lesion diagnostic systems for darker-skinned patients[11]. This highlights the need for external validation and diversity in clinical trials to reduce bias in digital health technology. 

Algorithms can also fail to account for social context mathematically. In 2019, a widely used healthcare resource allocation algorithm was found to deprioritize Black patients. The algorithm used healthcare spending to measure healthcare need. Because Black patients, on average, had lower healthcare spending due to reduced access, the model scored them as less sick than equally ill White patients[12].

Designing for Access, Usability, and Transparency

Representative clinical evidence is limited by the feasibility of participation. Digital biomarkers are one way we can increase the diversity of decentralized clinical trials. By reducing geographic and logistical barriers to access for underrepresented communities through home-based data collection, we can increase diversity in clinical trials. The FDA recognizes this, suggesting that remote approaches can improve the recruitment and retention of diverse participants[13]. 

However, the use of digital tools comes with important caveats. Broadband access, digital literacy, language, and device usability vary across the population. Decentralized trial designs must be evaluated for their impact on digital health equity. Evidence has shown that without usability testing across diverse subpopulations, tools may narrow rather than expand clinical diversity[14]. 

Data completeness across demographic subgroups is equally important. If dropout rates differ by demographic group, data completeness may become disproportionate, regardless of enrollment equity. Data completeness also enables subgroup analysis, which is important for discerning which subgroups may be more affected by a disease (e.g., colorectal cancer). However, evidence indicates that black participants are significantly less likely to use remote patient monitoring systems[15].

Transparency is the final crucial element. In the context of Black History Month, this is the most important consideration for minority communities. Diversity Action Plans, publicly reporting measurement performance across populations and disclosing enrollment equity metrics, are vital steps to rebuild the trust that has been eroded over decades of exclusion and exploitation. 

A Framework for Digital Health Equity has been proposed to identify digital determinants of health across societal levels, providing a structured method for identifying and removing barriers to equitable participation[16].

Equity as an Evidence Standard

A measurement that performs accurately for some participants but not others cannot be considered a reliable clinical endpoint. Diversity is therefore not separate from scientific quality; it defines it. Similarly, a dataset that represents some populations but not others limits confidence in scientific and clinical conclusions. The FDA’s work on Diversity Action Plans is an important start, ensuring that representative evidence becomes a vital standard of clinical research.

Sponsors, regulators, and measurement developers must design for equity across enrollment, analytic validation, and operational execution, ensuring that evidence generated reflects all intended populations with equal reliability. Delivering on this standard requires regulatory alignment, operational experience across diverse clinical programs, and the infrastructure to support consistent data quality across populations.

Designing clinical trials for equality at every stage will produce evidence that meets inclusivity standards and supports equitable patient outcomes across all underrepresented groups, particularly Black individuals.

Contact Koneksa to discuss how measurement strategy, analytic validation, and operational rigor can support inclusive, high-quality evidence generation across your clinical program.

Contact Koneksa

---

Frequently Asked Questions (FAQs)

What are FDA Diversity Action Plans?

FDA Diversity Action Plans are enrollment strategies that sponsors of Phase III and other pivotal clinical studies must submit under the Food and Drug Omnibus Reform Act of 2022. They must include specific enrollment goals, disaggregated by age, sex, race, and ethnicity, so that the clinical sample's demographics reflect those of the target patient population.

Why does device validation across skin tones matter?

Many digital health technologies, such as pulse oximeters and wrist-worn sensors, rely on light-based measurements. Melanin absorbs light differently across skin tones, which can affect measurement accuracy. If devices are validated primarily on lighter-skinned populations, clinical data collected from diverse participants may contain systematic inaccuracies. Validation across a representative range of skin tones, using frameworks such as the Monk Skin Tone scale, is essential to ensure that digital endpoints are reliable for all trial participants.

How does digital health technology affect clinical trial diversity?

Digital health technologies can both expand and limit the diversity of clinical trials. Remote data collection via wearable devices and home-based monitoring reduces geographic and logistical barriers, enabling broader participation by underrepresented communities. However, disparities in broadband access, digital literacy, and device usability can exclude the very populations these tools aim to include. Effective use of digital health in clinical trials requires usability testing across diverse populations and transparent reporting of data completeness by demographic subgroup.

---

References

1. Alegria M, Sud S, Steinberg BE, Gai N, Siddiqui A. Reporting of Participant Race, Sex, and Socioeconomic Status in Randomized Clinical Trials in General Medical Journals, 2015 vs 2019. JAMA Netw Open. 2021;4(5):e2111516. doi:10.1001/jamanetworkopen.2021.11516

2. Arias E, Xu J, Kochanek K. United States Life Tables, 2022. In: National Vital Statistics Reports [Internet]. National Center for Health Statistics (US); 2025. doi:10.15620/cdc/174575

3. Roope LSJ, Walsh J, Welland M, et al. Reducing inequalities through greater diversity in clinical trials – As important for medical devices as for drugs and therapeutics. Contemp Clin Trials Commun. 2025;45:101467. doi:10.1016/j.conctc.2025.101467

4. National Academies of Sciences E, Affairs P and G, Committee on Women in Science E, Research C on I the R of W and UM in CT and, Bibbins-Domingo K, Helman A. Barriers to Representation of Underrepresented and Excluded Populations in Clinical Research. In: Improving Representation in Clinical Trials and Research: Building Research Equity for Women and Underrepresented Groups. National Academies Press (US); 2022. Accessed February 18, 2026. https://www.ncbi.nlm.nih.gov/books/NBK584407/

5. Ndugga N. Racial and Ethnic Disparities in Access to Medical Advancements and Technologies. February 22, 2024. Accessed February 18, 2026. https://www.kff.org/racial-equity-and-health-policy/racial-and-ethnic-disparities-in-access-to-medical-advancements-and-technologies/

6. U.S. Food and Drug Administration. Digital Health Technologies for Remote Data Acquisition in Clinical Investigations. December 2023. Accessed February 20, 2026. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/digital-health-technologies-remote-data-acquisition-clinical-investigations

7. Al-Halawani R, Charlton PH, Qassem M, Kyriacou PA. A review of the effect of skin pigmentation on pulse oximeter accuracy. Physiol Meas. 2023;44(5):05TR01. doi:10.1088/1361-6579/acd51a

8. Sudat SEK, Wesson P, Rhoads KF, et al. Racial Disparities in Pulse Oximeter Device Inaccuracy and Estimated Clinical Impact on COVID-19 Treatment Course. Am J Epidemiol. 2023;192(5):703-713. doi:10.1093/aje/kwac164

9. Weir VR, Dempsey K, Gichoya JW, Rotemberg V, Wong AKI. A survey of skin tone assessment in prospective research. Npj Digit Med. 2024;7(1):191. doi:10.1038/s41746-024-01176-8

10. Koerber D, Khan S, Shamsheri T, Kirubarajan A, Mehta S. Accuracy of Heart Rate Measurement with Wrist-Worn Wearable Devices in Various Skin Tones: a Systematic Review. J Racial Ethn Health Disparities. 2023;10(6):2676-2684. doi:10.1007/s40615-022-01446-9

11. Tjiu JW, Lu CF. Equity and Generalizability of Artificial Intelligence for Skin-Lesion Diagnosis Using Clinical, Dermoscopic, and Smartphone Images: A Systematic Review and Meta-Analysis. Medicina (Mex). 2025;61(12):2186. doi:10.3390/medicina61122186

12. Obermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting racial bias in an algorithm used to manage the health of populations. Science. 2019;366(6464):447-453. doi:10.1126/science.aax2342

13. Research C for DE and. Conducting Clinical Trials With Decentralized Elements. October 16, 2025. Accessed February 19, 2026. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/conducting-clinical-trials-decentralized-elements

14. Aiyegbusi OL, Cruz Rivera S, Kamudoni P, et al. Recommendations to promote equity, diversity and inclusion in decentralized clinical trials. Nat Med. 2024;30(11):3075-3084. doi:10.1038/s41591-024-03323-w

15. Adepoju O, Dang P, Nguyen H, Mertz J. Equity in Digital Health: Assessing Access and Utilization of Remote Patient Monitoring, Medical Apps, and Wearables in Underserved Communities. Inq J Med Care Organ Provis Financ. 2024;61:00469580241271137. doi:10.1177/00469580241271137

16. Richardson S, Lawrence K, Schoenthaler AM, Mann D. A framework for digital health equity. Npj Digit Med. 2022;5(1):119. doi:10.1038/s41746-022-00663-0