Clinical Trial Site Feasibility: A Demonstration, Not a Questionnaire

At a glance

• ICH E6(R3) does not ask a site to estimate its recruitment. It asks the site to be able to demonstrate the potential, from retrospective or currently available data.
• The number that matters is the proposed number of eligible participants within the agreed recruitment period. Each of those qualifiers is where estimates usually break.
• Resources are three separate tests: sufficient time, an adequate number of available and qualified staff, adequate facilities. All three are qualified by the foreseen duration of the trial.
• The sponsor has its own obligation here, including giving sites the protocol and sufficient time to review it. A feasibility response returned in 48 hours is not an assessment.
• The regulations set no thresholds, no scoring, and no questionnaire content. That is a documentation problem, not a licence.

The requirement is evidentiary, and most feasibility is not

Site feasibility is usually run as a questionnaire. How many patients with this condition do you see annually, how many coordinators do you have, how many competing studies are you running. The answers come back, get scored, and the site is selected or not.

The requirement in the standard is a different kind of thing.

The investigator should be able to demonstrate, for example based on retrospective or currently available data, a potential for recruiting the proposed number of eligible participants within the recruitment period as agreed with the sponsor (ICH E6(R3) §2.2.1).

Read “demonstrate, based on retrospective or currently available data” against “answer a question about your patient volume” and the gap is obvious. One is an assertion. The other is a claim with evidence behind it. A site that answers the questionnaire confidently and cannot produce data supporting the answer has not met §2.2.1, no matter how thorough the form was.

This is not a call for more paperwork. It is a call for a different artifact. The useful feasibility record is not a completed questionnaire, it is a short statement of what the site’s own data shows about the eligible population, and where that data came from.

The three qualifiers people lose

§2.2.1 is a dense sentence and each qualifier does work.

“The proposed number.” Not a general sense of capacity. Feasibility is assessed against the enrolment target actually being asked of this site, which means feasibility cannot be assessed before that target exists. Sites are routinely asked to confirm feasibility for a study whose per-site target has not been set.

“Eligible participants.” This is the one that breaks estimates. Sites answer volume questions accurately and then under-enrol, because the question they answered was about patients with the condition, and the requirement is about patients who meet the protocol’s inclusion and exclusion criteria. A narrow biomarker requirement, a washout period, or a prior-therapy exclusion can remove most of a patient population, and a site cannot know that until it has read the protocol against its own records.

“Within the recruitment period as agreed with the sponsor.” Feasibility has a clock. A site that could enrol the target over three years and is being asked for one year is not feasible, and the questionnaire format tends to obscure this because it asks for annual volume without anchoring it to the study window.

Taken together, those three turn feasibility from “is this a busy site” into “can this site find this many of these patients in this time”, a question a questionnaire is poorly shaped to answer and site data usually can.

Resources: three tests, all over time

Recruitment potential is necessary and not sufficient. The second requirement is about whether the site can actually run the trial once those participants arrive.

The investigator should have sufficient time, an adequate number of available and qualified staff, and adequate facilities for the foreseen duration of the trial to conduct the trial properly and safely (ICH E6(R3) §2.2.2).

Three separate tests, and a qualifier that applies to all of them.

Sufficient time is about the investigator personally, and is the one most often waved through. An investigator with clinical commitments, three other trials, and a teaching load may be qualified and unavailable.

An adequate number of available and qualified staff. Note “available”, not merely “employed”. A coordinator shared across four studies is not fully available to any of them, and headcount answers on a questionnaire do not distinguish the two.

Adequate facilities. The most concrete test, and the easiest to check.

For the foreseen duration of the trial. This qualifier is what makes feasibility a forecast rather than a snapshot. A site that is adequately staffed today and is losing its only trained coordinator next quarter does not satisfy this, and no question about current headcount will surface it.

The sponsor-side mirror is stated as a selection duty: each investigator should be qualified by education, training and experience and should demonstrate they have adequate resources and facilities to properly conduct the trial (ICH E6(R3) §3.7.1). Again the verb is demonstrate.

The sponsor’s half of the obligation

Most feasibility content treats this as something done to sites. The standard puts a requirement on the sponsor too, and it is the one that quietly invalidates a great deal of feasibility practice.

The sponsor should provide the potential investigators and institutions with the protocol and an up-to-date Investigator’s Brochure, as well as sufficient time for the review of the protocol and the information provided (ICH E6(R3) §3.7.2).

Two obligations: give them the documents, and give them time to read them.

Set that against how feasibility usually runs. A questionnaire goes out with a synopsis rather than the protocol, a response is requested within days, and the site answers eligibility questions without having examined the actual criteria against its population. The site’s answer is then treated as a demonstration of recruitment potential, which under §2.2.1 it is not, and the sponsor’s own §3.7.2 obligation has not been met either.

Slow feasibility is not automatically better. But a feasibility process that structurally prevents a site from doing the analysis §2.2.1 requires cannot produce a compliant answer, only a fast one.

The capabilities a questionnaire rarely reaches

Two site obligations sit outside the usual feasibility form and determine whether a site can actually operate the trial it is being assessed for.

Whether the site can staff the work properly. The investigator should ensure that persons or parties to whom trial-related activities are delegated are appropriately qualified and adequately informed about relevant aspects of the protocol, the investigational products and their assigned trial activities (ICH E6(R3) §2.3.2), and should maintain a record of who those people are, proportionate to the significance of the activities delegated (ICH E6(R3) §2.3.3).

A site with headcount but no realistic path to getting those people trained on this protocol, and no habit of maintaining delegation records, will pass a staffing question and fail in conduct.

Whether the site can handle the product. Responsibility for investigational product management, including accountability, handling, dispensing, administration and return, rests with the investigator and institution, though the sponsor may facilitate aspects of it (ICH E6(R3) §2.10.1). Where that is delegated to a pharmacist or another individual in line with local requirements, the delegated person remains under the investigator’s oversight (ICH E6(R3) §2.10.2).

For trials with demanding product requirements, cold chain, controlled substances, complex reconstitution, this is frequently the binding constraint rather than recruitment. It is also where a generic feasibility questionnaire is least likely to ask a useful question, because the relevant detail is specific to the product and only becomes visible once the site has read the protocol, which returns to the sponsor’s §3.7.2 obligation to give them time to do so.

What feasibility is actually deciding

It is tempting to treat feasibility as a recruitment forecast with a compliance wrapper. The broader framing in ICH E8(R1) is more useful: the guideline addresses the consideration of quality in the design and conduct of clinical studies, including the identification, during study planning, of factors that are critical to the quality of the study (ICH E8(R1), objectives).

Site selection is part of study planning, and sites are one of the factors that determine whether the critical data can actually be produced. That reframes the question from “can this site enrol” to “can this site reliably produce the data that matters here”, which for a trial with a complex endpoint, a specialised assessment, or an unusual procedure may have a different answer than the enrolment question does.

A high-enrolling site that cannot perform the critical assessment consistently is a feasibility failure that a recruitment-focused questionnaire will score well.

What the standard does not say

Being honest about the limits matters, because the gap is where most feasibility disputes live.

The corpus sets no enrolment thresholds, no scoring methodology, no minimum staffing ratios, and no required questionnaire content. §2.2.1 and §2.2.2 state what must be demonstrated and leave the method to you. §3.7.1 makes selection the sponsor’s responsibility without specifying how to discharge it.

That is a real latitude, and it has a corresponding obligation: if the standard does not define the threshold, your own reasoning becomes the record. A feasibility decision that can be defended later is one where the sponsor can say what evidence it considered, what the site demonstrated, and why the conclusion followed, not one where a score crossed a line whose origin nobody can now explain.

The practical form of this is short. What did the site’s data show about the eligible population? Over what period, and against which criteria? What resource constraints were identified, and over what duration? What was concluded, and by whom? Four questions, answered in writing, will outperform a forty-item questionnaire in every conversation that matters later.

Reassessing, rather than assessing once

Feasibility is normally treated as a gate crossed once at selection. The resource requirement does not read that way. Sufficient time, adequate staff and adequate facilities are required for the foreseen duration of the trial (ICH E6(R3) §2.2.2), which is a condition that has to hold, not a state that has to be true on the day the questionnaire is returned.

That has a practical consequence sponsors tend to discover late. A site that was genuinely feasible at selection can stop being feasible, most often through staff turnover or a competing study opening in the same population. Nothing in the standard requires an annual re-questionnaire, but the obligation it describes is continuous, and sponsor oversight is where that gets noticed.

The cheapest form of this is not a process. It is knowing which two or three sites carry most of the enrolment, and treating a coordinator departure at one of them as a feasibility event rather than an administrative one.

The distinction worth keeping

Site feasibility asks whether a given site can run this trial. Protocol feasibility asks whether the protocol itself is runnable anywhere, and belongs before protocol lock rather than during site selection. They are different questions with different remedies, and the second has its own treatment.

Running the first as a substitute for the second is a common and expensive error: a protocol that no site can execute will produce a long list of infeasible sites, and the finding will be misread as a site-supply problem.

Sources

• ICH E6(R3) Good Clinical Practice, version R3
• ICH E8(R1) General Considerations for Clinical Studies, version R1