At the Mass General Research Institute, our community of 9,500+ investigators work diligently to publish peer-reviewed work and scientific findings to better understand disease and develop solutions to medicine’s most pressing challenges.

Footnotes in Science is a space where investigators bring you the behind-the-scenes details of their recently published work.

In this Q&A, we pick the brain of Brandon E. Turner, MD, MSc, the senior author of a new study published in American Academy of Pediatrics, Race and Ethnicity Reporting and Representation in Pediatric Clinical Trials

Dr. Turner is a resident physician in the Department of Radiation Oncology at Massachusetts General Hospital. In this study, he and his team sought to evaluate race and ethnicity reporting and representation in all US pediatric clinical trials on ClinicalTrials.gov

What motivated you and your team to publish this study?

It’s an exciting time to work in healthcare because of the many innovations across a wide array of fields and technologies, many of which are solidified through clinical trials.

Unfortunately, a vulnerability of our clinical research enterprise is the underrepresentation of important demographic groups.

This is a longstanding problem which has been really challenging to improve, but most of the evidence so far has been in adults.

We wanted to look at the pediatric population, as this group is particularly vulnerable and also demographically even more diverse than adults.

Can you touch upon the methods used to perform this study? 

We spent a lot of time developing tools to extract race and ethnicity information from trials in ClinicalTrials.gov.

Whereas most prior studies of trial diversity relied on manual review of isolated journals, guidelines, or studies, we were able to assemble a large dataset of over 1,000 clinical trials.

This allows us to avoid a lot of the noise resulting from small samples and to look for nuanced trends and relationships.

What are some key findings from your study?

Historically, most trials have not reported the race/ethnicity of their participants. However, this is actually getting much better (from <30% of trials in 2008 to >80% of trials in 2018).

Unsurprisingly, if you look in aggregate, all minority race/ethnicity groups are underrepresented relative to the population of minority children in this country.

However, because of the detailed data we collected, we were able to look at some specific disease areas (e.g. asthma, ADD/ADHD, autism, etc) and compare to this to epidemiological and hospital data.

When we take this lens, we see that the groups who are most consistently underrepresented appear to be Asian American and Hispanic children relative to their disease burden.

What criterion did your research team consider when selecting clinical trials?

We only included interventional trials conducted exclusively in the US. This is admittedly a limited lens in the modern era, as we increasingly rely on data from trials conducted all over the world when determining best practice.

However, in the US we do still practice primarily based on US evidence, which is reflective of the socioeconomic and policy environments in which our children experience healthcare.

Why is it crucial to have representative enrollment in biomedical research?

We have a lot of evidence that diseases and treatments vary from person to person, not everyone responds the same.

There are a lot of reasons for this, but one is that there are certain patterns that impact particular racial groups more than others.

While race is a social construct, there is often shared ancestry and heritable traits which can mediate the response to disease/therapy.

Unfortunately, in the US we also know that there are a lot of modifiable differences between groups, as far as access to healthcare, poverty, food insecurity, environmental exposures, etc.

For biomedical research to be objective, it has to sample from all of these groups. Otherwise we can end up with flawed understandings of diseases and produce therapies which only work for the portion of the population that we happened to include in the trial.

What are the next steps?

I really want to dig into the question of what is driving clinical trial diversities. Is this more of a structural process that will require systemic changes, or is it more individualized?

The latter in some ways is easier because it means each researcher and each trial is empowered to do better.

However, collective change is hard. If it’s more structural, our hands are tied a bit and we have to think bigger for how we can use big levers which move systems.

But either way we need to know, because a lot of people are working hard on this and we just want to make sure their efforts are being channeled in the most fruitful direction. 

What can research teams do to increase representative enrollment in their clinical trials?

This is a challenging question. There are incentives (positive or negative) that can be leveraged of course.

The FDA has recently tried to put pressure on industry sponsors of trials to come up with a sort of diversity enrollment plan prior to launching the trial.

I think if medical journals and societies took this into consideration as a factor in the quality and reliability of study results, I think that might encourage more academic sponsors to invest.

But I think the bigger challenge is that we still don’t have a great model for what is driving these diversities, and what are reliable strategies to improve them.

This is probably an area that needs more research, as right now there are a lot of promising ideas and anecdotal evidence, but not much high-quality mechanistic work.

The NIH and others could fund some studies which test novel approaches to study enrollment to actually test these ideas and really tease out the resource tradeoffs, as trialists don’t have infinite resources to recruit and inadequate accrual is already the most common reason that trials fail and are terminated.

What are some unanswered questions that remain from your study? 

I think, as above, we still don’t have a great mechanistic framework for what drives enrollment disparities and practical ways to fix it. I think we can do more to identify trials which have done a good job and learn from them.