Earning Trust in Communities Research Has Underserved

Most sites know their enrollment doesn't reflect the population they serve. The usual response is a push: a flyer campaign in a new ZIP code, a table at a health fair, a paid ad targeting a specific demographic. Then enrollment numbers barely move, and the conclusion becomes "that community isn't interested in research."

That conclusion is almost always wrong. What's usually missing is a reason to believe your site specifically will treat people well — and that belief is built slowly, through repeated, verifiable behavior.

Start by auditing what your site already signals

Before you plan outreach, look at your site the way a first-time visitor would.

• Who answers the phone, and in what languages? If a Spanish-speaking caller gets routed to voicemail, no amount of translated advertising will help.
• How long is the wait, and where do people sit? Sites in medical office parks can feel institutional in a way that reads as unwelcoming.
• Does your staff reflect your catchment area? Patients notice. Representation among coordinators and recruiters moves the needle more than most marketing tactics.
• Are consent documents actually readable? A 22-page form at a college reading level communicates that the study wasn't designed with ordinary people in mind.

Fix the things you control before you ask anyone to trust you.

Partner with institutions, not demographics

The sites that build durable, diverse enrollment tend to work through organizations that already hold trust: churches, barbershops and salons, federally qualified health centers, mutual aid groups, promotora networks, tribal health programs, immigrant service organizations.

Two rules make these partnerships work:

1. Show up before you need anything. Attend events for a year without a recruitment pitch. Volunteer at a screening day. Sponsor something. If your first contact is an ask, you're a vendor, not a neighbor.
2. Pay people for their time. Community advisory board members, interpreters, and partner organizations should be compensated like the professionals they are. Unpaid "collaboration" is extraction, and people recognize it.

Be direct about history

Many coordinators tiptoe around Tuskegee, around Henrietta Lacks, around forced sterilization, around the way immigration status affects willingness to sign anything official. Avoiding these topics doesn't make them absent from the room.

A better approach: name it plainly and then describe the specific protections that exist now — IRB oversight, the right to withdraw at any point without losing care, who can and cannot see their data, what happens if they get hurt. Skepticism rooted in real history deserves concrete answers, not reassurance.

Remove the logistics that filter people out

In most site datasets, the participants who drop out early aren't the least motivated — they're the ones with the least flexibility. Practical moves that matter:

• Offer evening and weekend visit windows, even a few per month
• Arrange and prepay transportation rather than reimbursing after the fact
• Pay stipends same-day where allowed, via a method that doesn't require a bank account
• Allow a family member in the consent conversation when the participant wants one
• Budget for interpretation in your feasibility numbers, not as an afterthought

Close the loop

The single most underused trust-building act in research is telling participants and partners what happened. Send a plain-language summary when results publish. Report back to the community organization that referred people. When a study fails, say so.

Communities that see results come back are communities that refer the next person. Communities that never hear anything again learn exactly what they were to you.