Recruiting for Studies Where Almost Nobody Qualifies
When the eligible population is a fraction of a percent, volume tactics stop working. Here's how to map protocol criteria for findability, mine your own records, and build the referral relationships narrow studies actually depend on.
By Trialflow Team
Narrow-criteria studies break the habits that work everywhere else. Broad advertising, high-volume call campaigns, and "get them in and we'll sort it out at screening" all collapse when the eligible population is a fraction of a percent of your database. The work shifts from generating volume to finding specific people.
Translate the protocol into a findability map
Before you advertise anything, go through inclusion and exclusion criteria and sort each one into three buckets:
- Findable in structured data — ICD codes, lab values, prescriptions, age, prior procedures. These you can query.
- Findable only in notes or by asking — symptom duration, prior therapy failure, functional status, caregiver availability.
- Determined at screening — washout-dependent labs, imaging, genotype.
The first bucket defines your candidate pool. The second defines your pre-screening script. The third defines your screen failure rate, which you should forecast out loud with the sponsor before you commit to enrollment numbers.
This exercise also exposes which single criterion is doing most of the filtering. In most narrow protocols one or two criteria eliminate the overwhelming majority of candidates. Build your search around that criterion first, then layer the rest.
Work your own records before you spend a dollar outside
For rare phenotypes, your chart data and prior study participants are usually a better yield source than any paid channel. Practical steps:
- Query for the single most restrictive structured criterion, ignoring everything else. Get the broadest defensible list.
- Manually review that list against bucket-two criteria. Yes, chart review by hand. For a list of 200 this is a day of work and it is cheaper than a month of ads.
- Cross-reference prior screen failures from related protocols — people who failed on one criterion may be clean for this one.
Track where each candidate drops out. After 30 or 40 reviews you will know which criterion is actually killing your funnel, and you can tell the sponsor something specific instead of "recruitment is slow."
Build referral relationships, not ad campaigns
When the condition is uncommon, the people who have it are already clustered somewhere — a specialty clinic, an infusion center, a patient advocacy group, a single subspecialist in your region. Those relationships out-produce broad media for narrow protocols.
Make referral easy and low-risk for the referring clinician:
- Give them a one-page criteria summary written in clinical shorthand, not protocol language.
- Give them a named person and a direct line, not a general number.
- Close the loop on every referral, including the ones you could not enroll, with the reason why. Referrers who never hear back stop referring.
Advocacy and condition-specific online communities are worth cultivating, but approach them as a long relationship rather than a campaign. Offer to answer questions about the disease and the trial landscape before you ask for anything.
Set expectations on the enrollment curve
Narrow studies rarely enroll linearly. You often get a cluster early from your chart review and known patients, then a long flat stretch where enrollment depends on new diagnoses walking through the door. Tell the sponsor that shape up front so the flat period is not read as site failure.
Finally, protect coordinator morale. Working ten candidates to enroll one is demoralizing if nobody names it. Measure effort in candidates reviewed and referral relationships maintained, not only in randomizations, and the team stays in the fight long enough to hit the number.
Enrolling studies shouldn't be this hard
Trialflow gives research sites one platform for lead management, AI trial matching, and recruitment analytics.
Request a Demo