Most site selection still runs on inferred data and a confident slide. Nobody bills you for the guess. The bill arrives later, itemized: an activation fee for a site that never enrolls, months of a flat curve, a rescue site at a premium, an amendment, and a program that slides. Here is what that invoice looks like, with the industry numbers behind each line.
Across roughly 16,000 sites in 151 Phase II and III trials, about 37% fell short of their enrollment target and about 11% enrolled no one at all. Each of those sites was chosen by someone who was confident. The feasibility number they were given was a promise, not a count.
Follow a site miss from feasibility to database lock. The costs do not arrive at once; they compound.
of patients who reach screening fail it, higher in oncology, CNS and rare disease. The pool quoted at feasibility is smaller than it looks before a single site opens.
sites never enroll a patient. The activation fee, the startup work and the monitoring plan are paid either way.
of trials miss their enrollment timeline. Recruitment is the single biggest driver of delay, and a flat curve at a weak site is where it starts.
median direct cost of a single substantial amendment, Phase II to Phase III. About 57% of protocols need one and nearly half are avoidable. Rescue sites get bolted on at the same moment, late and at a premium.
of enrolled patients drop out before the trial ends, and replacing each one costs multiples of the original recruit. A thin site never had the margin.
A single site that cannot reach the patients does not fail quietly. It sends five invoices, and each arrives after you can do anything cheap about it.
Startup budget, contracts, training and monitoring for a site that was never going to see a qualifying patient. The most avoidable dollar in the trial, spent first.
Months of a flat enrollment curve while the program clock runs, and while the strong sites you did not pick enroll for someone else.
New sites added late, at a premium, on a compressed startup, to catch up on enrollment that should have been in the plan from the start.
Eligibility rewritten because the pool turned out smaller than assumed. Enrollment resets against a moving target, at a median direct cost of $141K to $535K.
The whole program slides. For a CRO that is a sponsor relationship under strain. For a biotech it is runway, and the next raise moving further away.
Every line on the invoice traces back to one moment: a site chosen on a guess. Replace the guess with a count of the protocol-qualifying patients each investigator can actually reach, and the invoice never gets written.
Catch it before activation and you save the activation fee, the rescue site, the amendment, and the months.
You do not have to rip up a site list to de-risk it. Bring us the list you already have. We will tell you which sites will not enroll this protocol, and why, before you spend a dollar on activation.
Send us your site listIn a recent rare-disease program, counting the protocol against real, patient-level data nearly tripled the qualifying pool, and took the marquee centers with no qualifying patients off the list before activation dollars were spent.
We only publish figures we can point to. Where a number is an industry estimate rather than a published study, we say so.
What does guessing on clinical trial site selection cost?
A site that never enrolls still costs its activation fee, months of a flat enrollment curve, a rescue site bolted on late, often a protocol amendment at a median direct cost of $141K to $535K, and a timeline slip that pushes the whole program. One miss runs to hundreds of thousands of dollars and months of delay.
What percentage of clinical trial sites under-enroll?
Tufts CSDD analysis of roughly 16,000 sites across 151 Phase II and III trials found about 37% under-enroll and about 11% never enroll a single patient. Nearly half of activated sites fall short of what was promised at feasibility.
How much does a protocol amendment cost?
Tufts CSDD puts the median direct cost of a substantial amendment at about $141K for Phase II and $535K for Phase III. Around 57% of protocols need at least one, and nearly half of those amendments are considered avoidable.
How do you avoid rescue sites and amendments?
Catch the site that will fail before it is activated. Site selection built on real, patient-level data counts the protocol-qualifying patients each investigator can reach, so sites with no reachable patients come off the list before activation dollars are spent, and the enrollment curve behaves.
Bring us the protocol, or the site list you already have. We will show you which sites will enroll it and which will not, on real patient data, before activation.