Leveraging digital tools for more diverse trial recruitment
Digital recruitment tools get pitched fairly often as a straightforward diversity win: broader reach, lower barriers to finding out about a study, less dependence on which physical sites happen to be nearby. Recent work on digital tools for diversity and inclusion in recruitment takes that claim seriously and interrogates it properly, and the honest conclusion is more conditional than the pitch usually suggests.
Reach isn't the same as representative reach
A digital recruitment campaign can genuinely reach more people than a poster in a clinic waiting room. It doesn't automatically reach a more representative group of people. Social media advertising algorithms optimise for engagement, which can systematically favour certain demographics over others in ways that have nothing to do with a study's actual eligibility criteria. A campaign that looks broad in raw numbers can still under-reach exactly the populations a study most needs to include.
Where digital tools genuinely help, when built deliberately
The picture isn't uniformly discouraging. Several specific applications of digital tools show real promise, provided they're designed with representativeness as an explicit goal rather than an assumed side effect:
- Targeted outreach through community-specific digital channels, rather than broad, algorithm-optimised advertising alone, reaching populations through platforms and languages they actually use.
- Multilingual, culturally adapted digital materials, built from the start rather than translated as an afterthought once English-language materials are finalised.
- Digital pre-screening tools that reduce the burden of finding out whether someone's even eligible, which can particularly help potential participants for whom an in-person screening visit is a genuine barrier.
- Partnerships with community organisations to co-design digital outreach, rather than a study team assuming it knows how a given community prefers to be reached.
Where digital tools can quietly undermine diversity goals
The same review highlights failure modes worth watching for directly:
- Assuming digital access is universal. It isn't, and the populations without reliable digital access often overlap significantly with populations already underrepresented in research for other reasons.
- Treating a translated recruitment flyer as equivalent to genuine cultural adaptation. Language translation without adapting tone, framing, and trusted messengers for a specific community often falls flat.
- Relying on algorithmic ad targeting without auditing who it's actually reaching. A campaign can run successfully by engagement metrics while systematically underperforming with the specific populations a study needs.
The same tool, used two different ways
Several of the tools that help and the failure modes that undermine diversity goals are, confusingly, the same underlying technology used differently. Setting them side by side makes the difference in approach clearer than treating them as separate topics.
| Digital tool | Used to genuinely widen reach | Used in a way that quietly narrows it |
|---|---|---|
| Social media advertising | Targeted through community-specific channels, chosen deliberately | Left to an engagement-optimising algorithm with no audit of who it actually reaches |
| Translated materials | Culturally adapted, tone and messenger chosen for the specific community | A direct translation of the English-language flyer, nothing else changed |
| Digital pre-screening | Reduces the burden of an in-person visit just to check eligibility | Assumes every prospective participant has reliable access to complete it |
| Community partnerships | Communities co-design the outreach itself | A study team designs outreach alone, based on its own assumptions about a community |
The tool in each row isn't the variable that determines the outcome. The deliberateness behind how it's built and targeted is.
What auditing reach actually looks like in practice
"Audit who a campaign is actually reaching" is easy to state as a principle and vaguer to act on, so it's worth being specific about what this looks like operationally. A study running a social media recruitment campaign can typically pull basic demographic breakdowns from the advertising platform itself, age bands, broad location, sometimes language, and compare that against the demographic profile the study actually needs to enrol. Where platform data is too coarse, self-reported demographic questions at the screening stage, asked consistently regardless of which channel referred the participant, let a team see directly whether a given channel is bringing in a representative slice of interest or a narrow one.
The point of doing this mid-campaign, not just at final analysis, is that a channel underperforming on representativeness can still be corrected, reweighted, supplemented with a different channel, adjusted in targeting, while the study is still recruiting. Discovering the same gap only once enrolment is closed leaves nothing to do about it except note the limitation in the final report.
What this means practically for a recruitment strategy
- Set explicit representativeness targets before launching a digital campaign, not just an overall recruitment number, so under-reach in specific populations is visible early rather than discovered at analysis time.
- Audit who a digital campaign is actually reaching, not just how many people it reaches, using whatever demographic data is available from the platform or self-reported at screening.
- Combine digital outreach with non-digital channels deliberately, rather than treating digital as a full replacement for community-based, in-person recruitment approaches.
- Involve the communities a study needs to reach in designing the digital materials themselves, rather than assuming a study team's own instincts about messaging will translate.
The underlying point
Digital tools are a genuine opportunity to improve trial diversity, and dismissing them wouldn't be the right response to any of this. But they're a tool, not a guarantee. Whether a digital recruitment strategy actually widens who gets represented in a study's evidence base, or simply reaches more people who look like everyone the study was already going to enrol, depends entirely on whether representativeness was designed in from the start or assumed to follow automatically from reach.