Why Your Clinical Trial Isn't Enrolling (And What Actually Fixes It)

When enrollment stalls, the instinct is to blame the population: not enough eligible patients, not enough interest, the wrong ZIP code. In most stalled studies, the population is not the problem. The pipeline is, and it is usually failing in one of four specific, fixable places.

This guide walks through the four pipeline failures that actually stall enrollment at the site level, in the order a site should check them, and what closes each gap without waiting on a protocol amendment or a sponsor conversation.

The Diagnosis Most Sites Reach For First

A study that is behind on enrollment three months in gets read the same way almost everywhere: there must not be enough eligible patients nearby. It is the simplest explanation, and it is rarely the whole story.

In a meta-analysis of U.S. oncology trials, 56% of patients did not enroll because no suitable trial was available at their treatment location, 22% were ineligible under the study's own restrictive criteria, and 15% of eligible patients declined participation after being offered a slot (NCODA). Population scarcity is one line item in that breakdown, not the default explanation.

The guide on why sites miss their enrollment goals covers the full range of structural causes behind a slow study. Before writing off the local population, rule out the four pipeline problems below. They account for far more stalled enrollment than a genuine shortage of eligible patients ever does.

Your Eligibility Criteria Are Screening Out More Than the Protocol Requires

A protocol's written eligibility criteria and the way a site actually applies them at pre-screen are not always the same document. Coordinators under time pressure default to the stricter reading of an ambiguous criterion, which quietly narrows the eligible pool below what the protocol technically allows.

Three places this happens most often:

None of this shows up as a screen fail, because these patients never make it to a scheduled visit. They disappear from the pipeline as a soft no, which is exactly why the population looks smaller than it actually is.

Your Referral Pipeline Moves Too Slowly to Convert Interest

A patient who fills out an interest form is not yet a lost cause when the site takes two days to call back. But that patient is also not sitting idle. They are answering the phone for whichever study contacts them first, and speed to contact is one of the few enrollment levers a site controls completely.

A referral contacted within the first 10 minutes of applying is 4 times more likely to answer the screening call than one contacted a day later, based on Clinical Enroll client portal data across 30+ indications. Slower contact does not get logged as a decline. It just never converts, and the site never learns why.

A fast pipeline turns marginal interest into randomized patients. A slow one loses patients who were never counted as declined, just never followed up in time to matter.

Your Feasibility Numbers Were Never Checked Against the Real Population

The sites that hit their enrollment number on schedule ran the feasibility math before signing on, not after the study stalled. That means modeling how many patients in the local catchment area actually meet every criterion, not just the primary diagnosis, and comparing that number against the enrollment target the sponsor is asking for.

The guide on clinical trial feasibility assessment covers how to run that model against a specific protocol before committing to a number. Running it after three months of underenrollment tells a site what already went wrong. Running it before the contract is signed tells a site whether the number was ever realistic.

A free enrollment feasibility report runs this same modeling against a specific NCT number and ZIP code, a faster starting point for sites that have not built this check into their go or no-go process yet.

A Diagnostic Order Beats a Guess

Sites that fix enrollment fastest do not tackle all four problems at once. They check them in an order that isolates the actual cause before spending time or budget on a fix that was not needed.

Running the checks in this order takes an afternoon and usually surfaces the actual bottleneck well before the population gets blamed for it.

Two studies from the Clinical Enroll portfolio show what a working pipeline produces even against a genuinely narrow, hard-to-reach population: