Financial Factors Are Driving Inequities in Cancer Trial Participation

Clinical trials offer some of the most promising advances in cancer care, yet certain hurdles keep many patients on the sidelines. According to a new study appearing in the Journal of the National Comprehensive Cancer Network, these hurdles boil down to financial constraints.

Leveraging a unique dataset to examine these disparities at the patient level, investigators found that financial barriers, such as income, housing costs, and other assets, were the strongest predictors of participation in cancer clinical trials, even more so than factors like race and ethnicity.¹

In an interview with Targeted Oncology, Richard Hoehn, MD, assistant professor of Surgical Oncology at University Hospitals (UH) Cleveland Medical Center and study senior author, reviewed the study’s findings and outlined some of the structural changes needed to improve equitable access to clinical trials.

Targeted Oncology: What motivated you and your group to focus on this topic?

Richard Hoehn, MD: Cancer treatment equity and disparities in care have been a passion of mine for a while. It's my academic focus and my personal passion as far as studying and trying to ultimately fix these things. I have investigated with team members multiple different areas of cancer treatment disparities, and this was the first time we really looked at clinical trial inequities.

We had a unique opportunity: our hospital, our accountable care organization, had acquired this very robust LexisNexis socioeconomic health attributes dataset. It is patient-level socioeconomic characteristics, about 440 variables... Most studies that are published on socioeconomic disparities use address-based surrogates, so like the median income of a neighborhood or ZIP code, and that's almost all the studies that I've done. And that's…not as good as it can be.

[Now,] we have individual patient-level socioeconomic data for thousands of patients in our health system. We have a large regional cancer center that treats thousands of patients. We used that as an opportunity to say, “Let's look at all of our… thousands of patients [with cancer] over a few years, and then let's look at whether they participated in clinical trials, and use these patient-level data to understand that.”

I was very excited about this, because… all of the studies that I've read on clinical trials use datasets that are limited. Either they study only trial patients and describe them relative to the general population, or they use large datasets that have inaccurate trial participation information, or they use socioeconomic data based on addresses. Or really, a lot of the studies that have been done focus on age, gender, and race, because those…are more reliably collected. In any of the above study designs, there are real limitations, and so I thought that this would be a great opportunity to perform a large and granular analysis of a very important topic.

What were the most salient findings of your study, and did you find any of these findings surprising?

We started with over 400 patient level variables, and through a rigorous variable selection model that we've created over time… we ran multiple multivariate models and variable selection algorithms. Basically, all of the variables that were important were financial in nature: income, resources, cost of the home, cost of homes in the neighborhood, resources of the family and associates... All of the most powerful characteristics, far and away were finance related.

Things like health insurance are very, very important, obviously. Many health insurances prevent patients from enrolling in clinical trials. That's often the first thing we have to look at. Health insurance was used as part of our matching algorithm… Beyond finances and health insurance, [variables] like education, marital status, and race were either barely or not at all significant when you include all these financial characteristics.

I did not feel that this was very surprising. It is true that [factors] like gender have correlated with cancer disparities. Age certainly is a barrier to enrolling in certain clinical trials—older patients, in aggregate, tend to have more medical problems and be frailer. But at a patient level, these things are variable. Same with race; we found that when…we created a model only including race and ethnicity, that it was an important predictor. But when we introduced all these financial characteristics, race and ethnicity were no longer significant. What that tells me is that just your race or ethnicity does not necessarily predict if you're going to enroll in a clinical trial. But in Cleveland and northeast Ohio, certain racial and ethnic populations have different access to certain resources, and there's inequity across our country with regard to socioeconomic position and race and ethnicity. What it says to me is that it's not the race, the ethnicity, the gender, the age…generally, but it's your individual scenario. Are you able to take the time off work, come to multiple appointments? Do you have the support to go to the extra treatments and scans and labs and travel across town and do all the things necessary to participate in a clinical trial?

What efforts do you think need to be made at the structural level to mitigate some of these financial barriers?

I think it's a few things. First of all, there are Medicaid, Medicare Advantage, certain health insurance plans and policies that do not support clinical trial enrollment, and that is criminal in this day and age. The data are very clear that participating in a clinical trial leads to better quality care and better survival for patients [with cancer]. As a payer, if your job is to provide the best care you can for your covered patients, you should support clinical trial enrollment.

Secondly, we need to be smart about clinical trial design. Often, clinical trials in cancer are designed in a very rigorous way [for] these particular patients, who we know, who we have confidence, are going to tolerate treatment and jump through all of these hoops medically and physiologically, to participate in the trial. We want to be very rigorous in who we study, but what that does is it excludes a lot of patients. We make a lot of demands as far as scans, labs, and where you can have treatment. You often have to come to the main campus hospital to get treatment if you're going to enroll in a clinical trial, and all of these demands exclude patients.

If we want to be more equitable in who we enroll in clinical trials, I think clinical trial design has to be a little bit more open-minded and pragmatic in that patients can receive treatment throughout the region at multiple satellite locations. We allow monitoring, scans, labs…at a reasonable interval, and we don't place undue demands on patients. We don't make them go through a significant increase in testing compared [with] nontrial patients. I think those are probably the 2 biggest things: making sure that it's covered by health insurance and making sure trial design is reasonable and patient-friendly.

Your team is working on a larger study mapping “clinical trial deserts.” What characteristics define or determine these areas?

This paper has not been accepted for publication yet, but we're basically collecting data on clinical trial patients at the 3 major hospital systems in northeast Ohio and using that as the numerator, and the denominator being all cancer patients. Now, this will be stronger than the study we're discussing in that [it] is a much larger group of patients, and it represents the entire population of cancer patients in this 15-county region. But the limitation is that we don't have patient-level socioeconomic data for all of those patients like we do in our UH study, so it will be good that we can map what neighborhoods throughout the region have high and low clinical trial enrollment.

What we're seeing is that it follows the classic urban, suburban, rural “donut” phenomenon, whereby patients in the more affluent suburbs that ring around the major city have the highest trial participation in folks in the inner city and folks further out in the lower-income rural areas have lower trial participation. While we don't have the patient-level data, it follows a very similar socioeconomic pattern that we've seen in many other studies.

REFERENCE

1. Dong W, Shoag J, Lal T, et al. A novel evaluation of patient socioeconomic characteristics that predict clinical trial enrollment. J Natl Compr Cancer Netw. 2026;24(1). doi:10.6004/jnccn.2025.7092