The Systemic Flaw in Clinical Trials

Graeme M. Lipper, MD


October 09, 2020

This chaotic year got even more controversial this summer after the video-recorded death of George Floyd. The resultant nationwide protests calling for racial justice brought new attention to systemic racism. Sadly, that term has become unnecessarily controversial.

Efforts to address bias in medicine don't seek to point fingers or assign moral blame. Rather, identifying inequities in the quality and delivery of healthcare furthers the goal of us all to provide excellent care to patients from all racial, ethnic, and socioeconomic backgrounds. The objective of programs addressing this pervasive issue is simply to identify areas in which one racial or ethnic group is disadvantaged over others, and try to mitigate it.

Clinical trials are the bedrock upon which evidence-based medicine is built. When treating complex dermatologic conditions, we rely on data from well-designed, appropriately powered, randomized, placebo-controlled, double-blinded phase 3 clinical trials. And in order to be able to provide clinically useful answers for a diverse population, those trials have to reflect the real-world populations being treated.

In this light, a recent cross-sectional study examined whether trials of plaque-type psoriasis are flawed by intrinsic "ethnoracial disparities." The investigators gathered data from 82 completed phase 3 clinical psoriasis trials including almost 50,000 patients.

Most of these psoriasis studies were large, multinational trials with both public and private funding. Treatments included biologics (n = 11), topical agents (n = 10), and three oral agents (one phosphodiesterase type 4 inhibitor and two other oral systemic drugs).

The analysis yielded some startling results:

  1. The vast majority of participants (85.8%) in studies reporting ethnoracial data were classified as White.

  2. Black/African American participants accounted for only 3.09% of the combined study population.

  3. Only 76% of the 86 large phase 3 trials evaluated reported ethnoracial data. Only 65% of these reported comprehensive data. The remaining studies only reported the percentage of White participants or White participants plus one additional ethnoracial group.


Recently, a slew of articles in medical publications and the lay press have identified discrepancies in how dermatologists identify skin conditions in White persons vs persons of color. Suggested reasons include our education and training including inadequate exposure to skin of color during training and underrepresentation in dermatologic textbooks and journals. And once out in practice, many of us have less ongoing exposure to patients with skin of color owing to inequities in access to specialty care.

Combined, these factors result in people of color being less likely to receive accurate diagnoses and effective treatment for both common (eg, acne, psoriasis, atopic dermatitis, pityriasis rosea) and rare (eg, COVID-associated pseudopernio) skin conditions compared with their White peers.

This study adds the concern that phase 3 clinical trials are also subject to unintentional racial disparities. Using the example of plaque psoriasis, the authors identified two critical flaws: failure of one quarter of the trials to provide ethnoracial data, and underrepresentation of skin of color, especially Black/African Americans.

The first flaw is harder to understand. No well-designed clinical trial should omit comprehensive ethnoracial data, because it should be self-evident that not all populations respond to treatments in the same way. As one example, a multicenter cohort study recently showed that non-White ethnicity is a negative predictor of biologic therapy effectiveness in psoriasis.

The second flaw is more complex and probably due to several factors. Are fewer people of color enrolling in dermatology trials because of poorer access to specialty healthcare? What role does mistrust of clinical research play, given the legacy of unethical medical experimentation as exemplified by the infamous Tuskegee syphilis study?

We all agree that the results from a clinical trial in adults should not be applied to children. In a similar fashion, data derived from a sample population heavily skewed to one racial or ethnic group cannot be generalized to others. African Americans, for instance, have a higher burden of diabetes mellitus and genetic diseases, such as sickle cell anemia, that could drastically affect the safety and efficacy of some drugs compared with age- and gender-matched White patients. Interethnic differences in drug safety and response are well-documented. Just two examples:

  1. Primaquine-induced hemolysis is more frequent in Black persons than in White persons.

  2. Between 5% and 15% of ethnic Chinese patients express the HLAB*15:02 allele, compared with only 1% of White patients. This translates to a sixfold greater risk for carbamazepine-induced Stevens-Johnson syndrome/toxic epidermal necrolysis in Chinese patients, especially those with Han ancestry.

Our goal is to offer all of our patients the safest and most effective treatments for their skin diseases. To achieve this, the first step is to make accurate diagnoses regardless of age, gender, race, or ethnicity.

Equally important, we must ensure that the safety and efficacy of the treatments we offer have been thoroughly vetted in all racial and ethnic groups.

Graeme M. Lipper, MD, is a clinical assistant professor at the University of Vermont Medical College in Burlington, Vermont, and a partner at Advanced DermCare in Danbury, Connecticut.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.