The site activation list lands in the CTM’s inbox, and the enrollment diversity target sits in section 5.3 of the protocol. Forty-two sites selected, predominantly academic medical centers in urban research corridors, mostly the same sites that ran the last three studies. The sponsor’s diversity action plan promises 25% enrollment from Black and Hispanic participants. The screen failure report at week eight shows 11%. Nobody is surprised, and nobody changes the list.
A perspectives piece published this month in The Lancet makes a point that clinical operations teams should read as a direct indictment of standard site selection methodology. The authors document how the label “hard to reach” gets applied routinely to minority ethnic populations in biomedical recruitment, framing their under-representation as a community characteristic rather than a research infrastructure failure. Operationally, that framing has a concrete consequence: sponsors keep selecting sites that cannot reach those communities, then wonder why their diversity targets go unmet.
The label does the damage before the protocol is even written.
Where the Failure Actually Lives
Sponsors often treat diversity enrollment as a recruitment problem, something to be fixed in weeks ten through twenty when the demographic breakdown looks wrong. That is too late by months. The decision that determines whether a trial can enroll diverse participants happens at site selection, typically during feasibility, before a single IRB submission has been filed. A site that draws its patient population from a single ZIP code, staffs its study coordinators entirely in English, and has no existing relationship with community health workers cannot fix its diversity numbers by distributing flyers.
The BMJ Open study tracking racial and ethnic diversity in trials reported to ClinicalTrials.gov from 2009 to 2024 found that only 54% of studies reported both race and ethnicity data in 2024. Sponsors cannot manage what they do not measure, and for roughly half of all trials on the books, the demographic composition of enrolled participants remains invisible in the public record. That is not a data quality problem alone. It reflects how diversity has been treated operationally: aspirational language in the protocol, minimal infrastructure behind it.
The Clinical Trials Transformation Initiative’s Diversity Project, which began in 2019, frames the fix as requiring long-term, transformative strategies rooted in organizational commitment to building clinical trial research infrastructure that is more responsive to historically underrepresented populations. That is the right diagnosis. The operational translation is blunter: infrastructure means sites, and sites mean the feasibility questionnaire needs new questions.
Standard feasibility questions ask about investigator experience, patient volume, and competing protocols. Sites I work with across our network rarely get asked, at the feasibility stage, whether they have bilingual coordinators on staff, whether they conduct community outreach with local federally qualified health centers, or whether their IRB has experience with community advisory board engagement. Those omissions are not neutral. They systematically favor established research sites that already serve research-experienced, predominantly white patient populations.
What Purpose-Built Infrastructure Actually Produces
The counterargument from sponsors is predictable: diverse sites are slower to activate, carry higher regulatory risk, and have less experience with complex protocol requirements. That concern is real for sites with no prior trial experience. But it conflates two different categories. Community-affiliated sites with established patient relationships and trained coordinators are not inexperienced sites; they are sites that have been excluded from sponsor feasibility lists because their patient demographic did not match the historical enrollment target.
The evidence on what happens when site selection methodology actually changes is specific. Alliance Clinical’s published outcomes from an ongoing diabetic peripheral neuropathic pain study show that five of its purpose-built diversity-focused sites accounted for 37% of total enrollment across the entire study. Approximately 20% of participants enrolled across all sites were from racial and ethnic minority groups. Five sites, 37% of enrollment. That ratio does not happen by accident or by adding a flyer to the site’s waiting room. It comes from site selection criteria that weighted community embeddedness alongside the standard feasibility metrics.
Community-based participatory research methods show a similar pattern at a larger scale. A 10-year systematic review of CBPR approaches in clinical trial recruitment found that more than 85% of studies using these methods saw statistically positive outcomes in recruiting racial and ethnic minority participants. The mechanism is not mysterious: communities that have been involved in designing recruitment, that have relationships with the research team, and that are not being approached as passive subjects show up. The operational question is whether the sponsor’s site activation model creates the conditions for that participation to happen.
Turning the Compliance Box Into a Clinops Workplan
The Food and Drug Omnibus Reform Act of 2022 established the requirement for Diversity Action Plans for Phase 3 drug trials and certain device investigations, making demographic enrollment targets a regulatory obligation rather than a voluntary aspiration. Sponsors are now filing those plans. The gap between filing a plan and building the site infrastructure to execute it is where most programs are currently sitting.
The operational shift that closes that gap starts at feasibility. Add three questions to your standard feasibility template and weight the responses in site selection scoring: Does the site have bilingual coordinators for the target language population in the protocol’s geography? Does the site have an existing referral relationship with a community health organization, FQHC, or faith-based health program? Has the site enrolled at least one prior trial where 25% or more of participants were from a racial or ethnic minority group? Sites that cannot answer yes to any of those questions are not diverse-enrollment-capable regardless of their patient volume numbers, and activating them while expecting them to hit diversity sub-targets is the operational loop that produces the week-eight screen failure report.
For sites, the parallel move is documentation. If your site has community relationships, coordinator language capacity, or a track record of diverse enrollment, that evidence needs to be in your feasibility response, quantified and specific. Sponsors relying on terminology shifts away from “hard to reach” in their protocol language but not in their site selection criteria will keep selecting the same forty-two sites. Sites that can show enrollment demographics from prior studies, and name the community organizations they work with, give a sponsor the data to make a different decision.
The Lancet piece argues that the “hard to reach” label reflects a judgment about infrastructure, not about communities. Clinical operations teams have the lever to prove that. It lives in the feasibility questionnaire, and the next activation list is where it gets used or ignored.
References
- The Lancet, “A positive difference: unimagined communities, reimagined research”
- BMJ Open / PMC, “Racial and ethnic diversity in clinical studies reported to ClinicalTrials.gov, 2009–2024”
- Clinical Trials Transformation Initiative, CTTI Recommendations: Increasing Diversity in Clinical Trials
- Alliance Clinical, “A Practical Approach to Building Diversity into Clinical Trials Using Purpose-Built Sites”
- University of Kentucky Scholars, “Community-Based Participatory Research (CBPR) to Enhance Participation of Racial/Ethnic Minorities in Clinical Trials: A 10-Year Systematic Review”
- FDA, Diversity Action Plans: Food and Drug Omnibus Reform Act of 2022 (FDORA) Guidance
- PMC, Terminology shifts from “hard to reach” to “underrepresented populations” in clinical trial guidance
