A Bird’s-Eye View of Health Inequities in Clinical Trials: Observations, Lessons Learned, and the “Flight Path” Forward

Clinical Researcher—August 2021 (Volume 35, Issue 6)


Elizabeth Marshall, MD, MBA


Sometimes life seems to bring you full circle. Lately, I find myself having conversations about utilizing data to obtain insights with just as many people in the pharmaceutical industry—where I began my career path—as I do with those in healthcare. COVID 19 and its impact on the world population has taught us many hard lessons, but the one positive is inequity awareness.

Inequity awareness is an unfortunate theme that I have tried to both question and bring mindfulness to over much of my adult life. I am a physician, a researcher, a veteran, and a patient. I have provided insight and awareness around situations of health disparities for well over a decade. I have also tried to assist where I can.

I am happy to report that, now that this awareness has been brought into clear view, people seem to be much more receptive to learning about social determinants of health (SDoH). More importantly, this awareness has now prompted actions…or at least plans for future actions.

Bird’s-Eye View—Circling from Above

The odds of success for a Phase III clinical trial are 3 out of 5. Surely, those aren’t very good odds for such an expensive venture. The length of the Phase III trial also needs to be considered, because a lot can happen to those 300 to 3,000 research participants during the one to four years it takes to complete a trial. This is one reason that real-world evidence is imperative, and why we must take into consideration the SDoH information that is available in unstructured data.

Recent discussions I’ve had with folks in the pharmaceutical sector have certainly conjured many memories of the days when I was a researcher, including my observation that overall health is mainly influenced by factors outside the clinical setting. Capturing SDoH information as part of the clinical trial process can help patients, people, and the industry be more successful. Because so much can happen to influence the results beyond trial visits, clinical researchers can no longer accept business as usual.

Observations: Assessing the Need for a New Approach

Beware—and be vigilant of—unintentional bias

Bias may not be obvious, but it is there. Research cohorts are biased, not intentionally, but by the very nature of a study. Most participation requires hours of “free” time. Think about how many full-time working people that you know. Now—how many of them have been a participant in a clinical trial? Not many, I’m almost certain…especially if they need to sacrifice vacation time to do so.

With an unemployment rate regularly under 10% (thankfully so), not many people have the ability to participate with their work/family schedules. How does this biased collection of research cohorts represent the population? It doesn’t. Another unfortunate truth: there are those participants that partake in clinical trials as their career.

Social determinants of health are abundant

If you have ever had the misfortune of either suffering from or observing someone that is desperate for relief of an ailment, you aren’t alone. From my experience, those suffering are first in line for a new chance for relief. Pain and suffering affect so many aspects of one’s life. Many of the participants I recall had one or many SDoHs—they lived alone, lacked social support, had transportation issues, etc. Even if the participants didn’t enter the trial with said stressors, that doesn’t mean one or more didn’t subsequently develop.

I know plenty of people with an ailment who struggled maintaining a job or lost their social support over time. We know many clinical trials fail because participants drop out. By utilizing proper communication channels and understanding underlying issues, many missed appointments can either be avoided or addressed (e.g., rescheduled or replaced by a virtual visit). For example, addressing potential transportation problems with protocol changes may ameliorate problems with dropped participation.

Lessons Learned: Resetting Our Compass with NLP

The use of  natural language processing (NLP) to capture SDoH and other relevant data from visit reports provides an important means for identifying vital unstructured information that can be utilized in informed ways. We can say for certain that understanding SDoH can make an impact and improve clinical trial recruitment and results. Here are a few areas based on my own experience and important tidbits from colleagues:

  • Collect and address clinical trial recruitment inequities so that the trial cohort is characteristic of the wider population.
  • Determine how to translate observed environmental inequities into clinical research. (e.g., a higher rate of disease for those living in a particular environment, such as Vibrio vulnificus outbreaks after Hurricane Katrina).
  • Integrate and assess factors of health inequality within analyses, models, and initiatives.
  • Assess clinical trial outcomes with a more granular approach. For example, consider whether the drug itself failed or if there were extenuating circumstances that contributed to lack of efficacy.

The “Flight Path” Forward

Birds migrate because of their awareness that new resources are needed. We too should learn from such wisdom. It’s time to migrate away from our conventional thinking that we have more control than we actually have. In the end, neither clinical settings nor clinical trials have as much to do with our health portfolio as we may think. Our traditional approach to clinical care and clinical trials requires a new trajectory. We must embrace new resources and seek new understandings—and fly swiftly toward a path that is more beneficial for all.

Elizabeth Marshall, MD, MBA, is Director of Clinical Analytics at Linguamatics, an IQVIA company.