A small-to-midsize biotech sponsor was preparing to initiate a Phase II clinical trial with timelines that, on paper, looked like most others in the industry: optimistic, and quietly at risk. Unwritten Health was engaged before protocol lock. What followed compressed a 9–12 month start-up to a projected 6 months.
Clinical trials routinely fail to reflect the populations they are intended to serve. Underrepresented groups, including people from lower socioeconomic backgrounds, ethnic minorities, and those living in areas with limited healthcare infrastructure, are consistently excluded, not through intentional discrimination, but through structural design choices made without community insight.
The consequences extend beyond health equity. Overrestrictive criteria shrink the pool of eligible patients, slow recruitment, and increase the risk of costly protocol amendments later in the trial. Research shows that over half of all trials submit at least one protocol amendment, with changes to trial population and eligibility criteria among the most common. Every avoidable amendment represents lost time, lost money, and delayed access to treatment for patients.[1,2]
The single most important factor in what followed was timing. Unwritten Health was brought in before protocol lock, the point at which study design is formally fixed and changes become expensive to make. This is a critical window that most sponsors do not use effectively.
Unwritten Health's Equity Engine platform draws on longitudinal, community-generated data from underserved populations, capturing not just clinical characteristics but the social determinants of health that shape whether someone can realistically participate in a trial. Three areas of intervention made the material difference.
Each eligibility criterion was examined against how people in the target communities actually live. Criteria that appear scientifically neutral in a protocol document can effectively rule out entire demographic groups in practice: age cut-offs that exclude older adults, laboratory value thresholds that do not account for conditions more prevalent in minority populations, and comorbidity exclusions that reflect the health profile of wealthier, better-served patient groups. By reviewing the proposed criteria through a community lens and flagging those that would exclude more patients than the science required, the protocol was refined before those exclusions were locked in.
The visit schedule was reviewed against real-world constraints that sponsors rarely account for. For working people in part-time or shift-based employment, frequent clinic visits mean loss of income. For carers, who are disproportionately women and disproportionately from lower-income communities, visits without flexible windows represent a structural barrier. Where the protocol was creating avoidable drop-off risk, adjustments were proposed before the design was fixed. Trials that reduce patient burden before enrolment begins achieve materially better retention.
The original strategy leaned heavily on hospital and academic medical centre sites, the defaults for most sponsors, but also the sites that underserved communities are least likely to access, least likely to trust, and most likely to find geographically distant. WCG Clinical data shows that median activation time at academic medical centres is 8.12 months, compared to 4.37 months at independent community sites, a difference of nearly four months before a single patient has been enrolled.[3] The site strategy was revised to include a community mix alongside the hospital network, driven by community data rather than post-hoc course correction.
The combined effect of these three interventions is a projected study start of six months, against a typical benchmark of nine to twelve months. That is two to four months recovered before the study has enrolled its first patient.
The industry has known about diversity and inclusion challenges in clinical trials for decades. Yet the same problems recur: underenrolment of minority populations, protocol amendments driven by unfeasible criteria, and retention failures that could have been anticipated.
The difference with a community intelligence approach is that it operates upstream, with data, rather than downstream with aspiration. Community engagement is not a communications strategy applied after the protocol is locked. It is a design input applied before the protocol is locked, using longitudinal data about how people in these communities actually live, what their real health profiles look like, and what barriers they face that a standard feasibility exercise will never surface.
The biotech sponsor in this case study did not receive a diversity report. They received a protocol that is more likely to reach its recruitment target, less likely to require an expensive amendment, and positioned to start six months sooner than the industry average. That is what community intelligence actually delivers.
The framework behind this approach is set out in full in our white paper, "Health equity risk in clinical research: how to see patient recruitment failure before it happens" — free to download alongside our other research on the Resources page.
Unwritten Health works with pharma, biotech, and CRO teams in the design phase, before timelines are at risk. Book a call to discuss your pipeline.
© 2026 Unwritten Health. All financial estimates are based on publicly available, peer-reviewed industry data. Clinical timeline and cost data are sourced from Tufts CSDD (2024) and WCG Clinical (2024). Individual study outcomes will vary.