Recruitment Assumptions and Where They Break

Part of Clinical Project Management (PM) Training: Building a Timeline That Holds.

About 2 minutes. Free, in your language, no account needed.

Every recruitment plan contains an assumed rate

Somewhere behind the enrollment curve is a number: participants per site per month. It came from feasibility questionnaires, a previous trial, or an estimate. It is the single assumption most likely to be wrong, and the one that moves the end date most.

Why sites over-predict

Not dishonesty. A site counts the patients with the condition, not the patients who also meet every inclusion and exclusion criterion, are willing to be randomized, can attend the visit schedule, and are not already in another study.

The gap between those two numbers is usually large, and nobody at the site has a reason to compute it during feasibility.

The early signals

Watch the funnel, not the total. Screened, screen failed, randomized. A site that screens many and randomizes few has an eligibility problem. A site that screens nobody has an access or a staffing problem. These need different responses, and the enrollment total alone cannot tell them apart.

The first eight weeks after activation tell you more than the projection did. A site that has not randomized anyone by then rarely recovers to its predicted rate without intervention.

Responses that work and one that does not

Adding sites is slow and expensive; each new site repeats the whole activation chain. Widening eligibility requires an amendment and changes what the trial can conclude. Supporting an existing site with a coordinator or referral pathway is faster than both.

The response that does not work is asking a site to try harder. If the constraint is eligibility or staffing, encouragement changes nothing and costs you credibility with the site.