Picture a Danish adult, somewhere in their sixties, sitting at a kitchen table and watching a short informational video about an RSV vaccine trial. The science in the video is identical regardless of which version they receive. The consent language is identical. The risk disclosures are identical. The only variable is who is looking back at them from the screen. A man or a woman. And that single variable, according to a study published in JAMA embedded within the DAN-RSV adult respiratory syncytial virus vaccine trial in Denmark, is associated with whether that person actually enrolls.
Most trial operations teams have never written a line in their recruitment protocol about the demographic presentation of their enrollment media. They have agonized over site selection, IRB timelines, inclusion criteria, and screen failure rates. But the face in the video? That has been, functionally, an afterthought. The DAN-RSV nested trial suggests that neglect carries a measurable cost.
The Mechanism Nobody Budgeted For
The behavioral science underneath this finding has a name: homophily. First formalized in sociology and later applied rigorously to health contexts, homophily describes the human tendency to trust, follow, and affiliate with people who resemble us along visible demographic dimensions. A 2018 randomized multi-center study published by Berry, Halpenny, and McMenamy examined homophily’s influence on health decision-making across 293 men, and found that demographic similarity between the information presenter and the recipient measurably shifted decision-making outcomes. The DAN-RSV video trial is, in effect, a direct test of whether that laboratory finding holds in a real-world vaccine trial setting.
Here is how the mechanism works, step by step. When a potential participant encounters an informational video, they are not processing a neutral data transmission. They are conducting a rapid, largely unconscious social appraisal: does this person know what I know about my own body? Do they share the life experience that made me hesitant about vaccines, or about clinical trials, in the first place? The presenter’s sex operates as a heuristic proxy for that experiential overlap. It does not determine enrollment on its own, but it shapes the emotional receptivity that precedes a rational decision.
This is where most protocol designers make a critical error. They assume that a well-designed consent video is a factual delivery mechanism. Regulators have historically treated it the same way: the FDA’s guidance on informed consent focuses on comprehension, voluntariness, and disclosure, but says nothing about the social encoding embedded in who delivers the information. The ICH E6(R3) Good Clinical Practice guidance, formally adopted by the FDA on September 8, 2025, advances the field considerably on participant protections and risk-based monitoring, but it still treats the recruitment media as a communication artifact rather than a behavioral intervention in its own right.
That framing creates a structural blind spot. If a sponsor optimizes their protocol for statistical power based on historical enrollment rates, and those historical rates were themselves shaped by demographic mismatches between presenter and target population, then the power calculation is built on a compromised baseline. The projected enrollment velocity was always lower than it could have been, and nobody flagged it because nobody was measuring presenter effects as a variable.
The DAN-RSV trial corrected for that by making the presenter’s sex the explicit randomization variable. This is methodologically elegant: rather than treating video production as a production decision, the investigators treated it as a scientific question. The result is a piece of evidence that most sponsors are not equipped to act on, because their protocols contain no mechanism for capturing or responding to it.
A Contrasting Signal From the Field
The DAN-RSV finding does not exist in isolation. The bivalent RSVpreF vaccine effectiveness data from the broader DAN-RSV program reflects a pragmatic, population-level trial design built around real-world conditions rather than controlled clinical center environments. That pragmatic framing is exactly what makes the nested presenter-sex study so consequential: it demonstrates that behavioral variables can be randomized and measured even within a large, community-facing trial architecture. The infrastructure was already built for heterogeneity. The investigators simply directed some of that heterogeneity toward an answerable question about human behavior.
Consider what that implies for every decentralized or hybrid trial running today, where informational videos have replaced the in-person recruiter visit. In a site-based trial from fifteen years ago, the “presenter” was whoever staffed the clinic that day. Variation existed but was unmeasured. In a DCT, that variation has been collapsed into a single produced artifact that plays identically for every potential participant in the arm. Sponsors have traded unmeasured demographic variation for locked-in demographic uniformity, and most have not asked whether the uniform choice they made was the right one for their target population.
What Skeptics Get Wrong About “Minor” Protocol Tweaks
The standard critique of this kind of finding goes roughly as follows: recruitment optimization studies are fishing expeditions. You randomize enough minor variables and eventually something crosses a significance threshold by chance. The cost of acting on spurious recruitment signals is protocol instability, IRB amendment burden, and noise in your enrollment projections. Better to standardize and move on.
That critique has operational merit, but it concedes the wrong premise. The presenter-sex variable in DAN-RSV was not a mining exercise. It was a pre-specified, randomized test of a theoretically grounded behavioral hypothesis. The homophily literature gave investigators a mechanism before they tested it. That is exactly the evidentiary standard the field applies to clinical endpoints, and there is no principled reason to apply a lower standard to recruitment endpoints when recruitment failure is the single most common cause of trial delay and budget overrun. A 2018 analysis across the clinical trials literature consistently identified enrollment shortfalls as the primary driver of trial extension costs, which regularly add months and tens of millions of dollars to Phase 3 programs.
The regulatory parallel deserves attention here. Sponsors submit protocol amendments for dosing adjustments, endpoint refinements, and eligibility criterion expansions, all of which require formal justification and, in many cases, revised statistical plans. But swapping the face in an enrollment video? No amendment. No power recalculation. No regulatory record. The ICH E6(R3) framework, for all its modernization around risk-based quality management, contains no explicit requirement that sponsors prospectively validate their recruitment media as behavioral interventions. That gap means the DAN-RSV finding, however robust, lands in a regulatory vacuum. Sponsors can act on it or ignore it with equal impunity.
The deeper problem this surfaces involves how the field categorizes what counts as a protocol variable. Statistical assumptions about enrollment rates flow through to sample size calculations, which flow through to interim analysis triggers, which flow through to the Data Safety Monitoring Board’s stopping rules. Every downstream decision in the trial architecture rests on enrollment assumptions that were built without accounting for behavioral effects that can now be measured. The DAN-RSV nested study has handed the field a tool, and most sponsors are not structured to pick it up.
Recruitment media should be treated as a testable intervention, not a production deliverable. The question sponsors should be asking before they lock a video is the same question they ask before they lock a dose: what is the evidence that this is the right choice for this population, and what would we do if we were wrong?
The face in the frame has always been doing work. The DAN-RSV trial is the first time many in the field will be forced to account for it.
References
- JAMA — “Effect of Information-Video Presenter Sex on Recruitment” (DAN-RSV nested trial)
- PMC / Berry, Halpenny, McMenamy — “Understanding Health Decision Making: An Exploration of Homophily” (2018)
- Diabetes Care / DAN-RSV — “Bivalent RSV Prefusion F Protein-Based Vaccine Effectiveness” data
- FDA — ICH E6(R3) Good Clinical Practice Guidance (adopted September 8, 2025)
Moe Alsumidaie is Chief Editor of The Clinical Trial Vanguard. Moe holds decades of experience in the clinical trials industry. Moe also serves as Head of Research at CliniBiz and Chief Data Scientist at Annex Clinical Corporation.
