The call nobody wants to chair happens around month four of enrollment: the sponsor’s clinical operations lead pulls together the CTM, the site liaison, and the data manager to explain why the diversity targets written into the protocol are not being met. The IRT stratification was set at activation. The sites were selected six months ago. The ICF was translated into two languages. And still, the enrolled population looks nothing like the one the Diversity Action Plan described. At that point, the operational options range from expensive to very expensive, and the amendment clock has already started.
A pilot program examining early-stage inclusion planning found that sponsors who build diversity strategy into protocol design, rather than grafting it onto an enrollment plan after sites are already activated, are better positioned to meet inclusion targets without mid-trial course corrections. That finding sounds obvious until you map it against how most sponsor organizations actually sequence their work.
Where the Sequence Breaks Down
In practice, diversity and inclusion planning at most sponsors sits downstream of protocol finalization. The clinical team writes the eligibility criteria. The biostatistics group sets the sample size. Then, once the protocol is locked, the clinical operations team is handed enrollment targets that implicitly require a patient population the site network was never specifically selected to reach. The Diversity Action Plan, now mandated under the Food and Drug Omnibus Reform Act of 2022, gets drafted as a companion document rather than a design input.
The operational consequences hit sites first. A coordinator at a site that serves a predominantly white, insured, employed patient population cannot manufacture a diverse enrolled cohort by working harder. No amount of additional outreach effort or extended screening windows changes the fundamental demographics of a site’s catchment area. If the sponsor did not select for community health centers, federally qualified health centers, or academic medical centers with documented diversity track records during feasibility, that gap cannot be closed after the SIV.
Sites I work with across our network report a consistent pattern: diversity-related protocol amendments often arrive in the early months of enrollment, after screen failure data has made the gap visible. By that point, the site has already built its patient identification SOP around the original eligibility criteria, trained coordinators on the original ICF, and set community outreach cadences that reflect the original target population. An amendment that expands eligibility or adds a recruitment stratum requires retraining, IRB resubmission, ICF versioning, and in many cases an IRT configuration change to add or rebalance stratification factors. The IRB turnaround for a substantial amendment, compounded by the IRT vendor’s configuration lead time for a stratification change, can mean the site loses weeks of productive enrollment while the paperwork catches up.
The economics are worse than the timeline. A site running at 1.5 patients enrolled per month on a rare disease protocol cannot absorb a two-month pause without the PI redirecting attention to a competing study. Coordinator overtime during the amendment scramble is rarely covered by the existing site budget, and the renegotiation of per-patient fees to reflect expanded screening work almost always lags the amendment itself by another billing cycle. The NIH’s inclusion policy for women and minorities has required demographic enrollment tracking for NIH-funded research for years, yet commercial sponsors are still discovering mid-trial that their site networks were not built to deliver on comparable commitments.
What Early Planning Actually Requires
Moving diversity planning upstream means treating it as a site selection criterion, not a recruitment tactic. The difference matters operationally.
A site selection criterion changes the feasibility questionnaire. It means asking prospective sites not just about total patient volume but about documented enrollment of specific demographic groups in prior studies, relationships with community health workers or patient navigators, language capability of the clinical team, transportation support infrastructure, and whether the site has experience running translated ICF versions without coordinator error rates that inflate deviation logs. These are answerable questions during PSV preparation. They almost never appear on standard feasibility surveys.
IRT and RTSM setup is where early planning has the clearest downstream payoff. Stratification factors for race, ethnicity, or other demographic variables must be built into the randomization algorithm at system build, before first-patient-in. Adding a stratum mid-trial requires a protocol amendment, a system change order, potential unblinding risk review, and a validation cycle the IRT vendor will bill against. Sponsors who define their diversity stratification targets during protocol design can hand those parameters to the IRT team during the system specification phase, when configuration changes cost a remark in the build log rather than a formal change order. The FDA’s guidance framework for Diversity Action Plans anticipates this kind of prospective planning; the pilot’s findings suggest that sponsors who act on it operationally, rather than just documenting it, avoid the amendment cycle that otherwise follows.
Patient identification SOPs at the site level also need to be written against the final enrollment target, not revised toward it. A coordinator building a screening log template, a referral tracking sheet, or a community outreach calendar does that work once if the protocol is clear at activation, and three times if the target population shifts mid-study. Across a large multi-site trial, repeated rounds of SOP revision at each site are not a paperwork inconvenience; they constitute a training event, a documentation audit, and a potential source of protocol deviations if any site implements the revision inconsistently.
The Monday Morning Directive
For sponsor clinical operations leads, the actionable change is to put inclusion planning on the protocol development timeline, not the clinical operations startup timeline. That means the feasibility team benefits from demographic performance data from prior studies when finalizing the site list, and the IRT specifications should reflect diversity stratification parameters as early as possible in the system build process. Neither of those steps requires new technology or new headcount; they require sequencing decisions made earlier in the planning calendar.
For sites, the ask is simpler but equally concrete: build your patient identification SOP to explicitly include the sponsor’s DAP-defined target population, and document your community referral pathways during SIV preparation rather than inventing them after the first enrollment review call. If the sponsor has not shared their inclusion targets by the time you are building your site-level recruitment plan, ask for them in writing before the SIV. A site that cannot produce those documents during sponsor oversight activities is poorly positioned if FDA ever reviews the enrollment record against the DAP commitments.
The sponsors who treat Diversity Action Plans as a regulatory filing obligation rather than an operational design input will keep chairing that month-four call. The ones who move inclusion planning into the protocol development phase will find that their site networks, their IRT configurations, and their coordinators’ SOPs are already aligned when enrollment opens, which means the diversity target is a deliverable, not a crisis.
References
- Clinical Trials Arena, “Developing early inclusion plan for trials benefits commercial sponsors, pilot finds”
- Applied Clinical Trials Online, “An Overview of FDA Diversity-Related Documents for Clinical Trials” (Food and Drug Omnibus Reform Act of 2022, Diversity Action Plan mandate)
- FDA, Diversity Action Plan Guidance Framework
- NIH, Policy and Guidelines on the Inclusion of Women and Minorities as Subjects in Clinical Research
