County-Level Trial Access Did Not Track Cleanly With Race and Ethnicity
Geography shaped access to uterine cancer clinical trials in ways that were not captured by a simple yes-or-no count of whether a county had any study site at all. In a retrospective cohort of 155,811 adults diagnosed with uterine cancer between 2008 and 2021, investigators linked cancer registry data, ClinicalTrials.gov trial listings, and county-level community data to ask a direct question: where were patients living when trials were available, and who was more likely to live near them?
The answer was uneven. Among the full cohort, 121,525 patients, or 78.0%, lived in a county with at least one available interventional uterine cancer trial. But the pattern varied sharply by race and ethnicity. Trial availability was observed for 58.0% of American Indian or Alaska Native patients, 94.3% of Asian patients, 77.7% of Black patients, 86.1% of Hispanic patients, 90.3% of Native Hawaiian or Pacific Islander patients, and 74.1% of White patients. Those differences are not just statistical artifacts; they reflect where trial infrastructure is concentrated and where it is thin on the ground.
A Single Trial in a County Did Not Mean Comparable Access
A county that contains one trial site can still leave most patients without a realistic pathway into research participation. Rural distance, transportation burden, trial eligibility rules, referral patterns, and language barriers all affect whether geographic presence becomes actual access. The study’s design made that distinction visible by separating the odds of living in any county with a trial from the density of trials available among patients already living in those counties.
Compared with White patients, Asian patients had much higher odds of residing in a county with at least one available trial, with an odds ratio of 5.83 (95% CI, 4.83-7.12). Black patients also had higher odds, with an odds ratio of 1.18 (95% CI, 1.07-1.30), as did Hispanic patients, with an odds ratio of 2.18 (95% CI, 1.80-2.63), and Native Hawaiian or Pacific Islander patients, with an odds ratio of 3.34 (95% CI, 2.17-5.66). American Indian or Alaska Native patients stood apart in the opposite direction, with lower odds of living in a county that had at least one trial, at 0.49 (95% CI, 0.39-0.61).
That first analytic layer points to broad geographic inequity. It also shows why access metrics need more than a presence count. A county with one trial is not equivalent to a county with several actively recruiting studies, especially for patients with aggressive or recurrent disease who may need a narrower eligibility window.
Trial Density Told a Different Story Than Trial Presence
Once the analysis shifted to patients living in counties where at least one trial existed, the pattern changed. Asian patients had fewer available trials per person than White patients, with a rate ratio of 0.69 (95% CI, 0.65-0.74). Hispanic patients had an even lower rate ratio of 0.56 (95% CI, 0.52-0.61), and Native Hawaiian or Pacific Islander patients had a rate ratio of 0.87 (95% CI, 0.80-0.95). In other words, some groups were more likely to live in a county with a trial at all, but once there, they still encountered fewer options.
This split matters clinically. Trial access depends on the number of nearby studies, not merely the presence of one local protocol. A county with sparse trial density can still force patients into long waits, closed enrollment windows, or mismatched eligibility criteria. For uterine cancer, where biologic subtype, stage, prior therapy, and biomarker requirements often determine trial eligibility, a thin local portfolio can effectively shrink the treatment-research interface.
Why Geography Becomes a Research Barrier in Uterine Cancer
Uterine cancer care often unfolds across multiple decision points: surgery, adjuvant therapy, recurrence management, and molecular testing. Clinical trials enter that pathway when disease characteristics match a study and when the patient can physically reach the site in time. Counties with concentrated academic and specialty oncology centers are more likely to host trials, and those centers are not evenly distributed across the U.S. population.
The county-level approach used here is especially useful because it reflects how patients actually experience access. Patients do not enroll in a national trial ecosystem in the abstract; they enroll from where they live. If a trial exists in a county but the density is low, patients may still face a practical shortage of options. If no trial exists nearby, the barrier becomes immediate and structural.
This is particularly relevant for populations already burdened by disparities in cancer outcomes and care access. Geographic trial scarcity can compound differences in referral networks, insurance coverage, availability of subspecialists, and the logistical support needed for participation. For American Indian or Alaska Native patients, the lower odds of county-level trial presence may point to a gap that is more basic than enrollment preference: the studies may simply not be nearby.
What the Cohort Design Clarified About Access Measurement
The analysis drew on adult patients diagnosed from January 1, 2008, through December 31, 2021, and used hurdle models to separate two questions: whether a patient lived in a county with at least one uterine cancer trial, and how many trials were available per person among those already living in counties with trial availability. Models accounted for county clustering and adjusted for patient characteristics.
That structure is methodologically important because it avoids collapsing two very different access problems into one endpoint. Presence and density can move in opposite directions, as they did here. A county may cross the threshold from zero to one trial, yet still offer limited practical reach. For a disease like uterine cancer, where trial participation can influence access to novel systemic therapy, targeted agents, and biomarker-driven treatment approaches, the number of available studies can shape care long before enrollment is discussed.
The median age at diagnosis was 62 years, with an interquartile range of 55 to 69 years. The cohort included 100,348 White patients, 23,177 Hispanic patients, 16,303 Black patients, 12,753 Asian patients, 2,175 Native Hawaiian or Pacific Islander patients, and 1,055 American Indian or Alaska Native patients. The distribution itself reminds readers that some groups remained numerically small, which can limit precision even when the direction of effect is clear.
The Endpoint That Matters Is Not Just Presence, But Reach
The study’s central message is straightforward: measuring whether a county has any uterine cancer trial can obscure the more consequential question of how many trial opportunities a patient actually has. Geographic access is not binary. It is layered, and the layers do not line up evenly across racial and ethnic groups.
For trial designers, cooperative groups, and oncology systems, the practical implication is to think beyond site counting. Trial placement, referral routing, and recruitment support need to account for county-level variation in both presence and density. Without that, a map with trial dots can look healthier than the patient experience it represents.
The study stops at geography and does not measure enrollment itself, so the link between county availability and actual participation remains unquantified here. Even so, the access pattern is concrete enough to raise a hard operational question for future trial planning: where are the studies, how many are there, and which communities are still left with too few choices?