Clinical Trial Site Feasibility Analysis: How Sponsors Evaluate Sites

When a sponsor or CRO has a protocol to place, they don't pick sites on reputation alone. They run a feasibility analysis: a structured comparison of candidate sites against what the study needs. Your questionnaire answers are the raw material. This article is about what the analysis weighs, how it's typically scored, and how to run it on yourself before a sponsor does — because the sites that get selected tend to be the ones that already knew their answer.

If you're new to the process, start with what site feasibility is and the feasibility questionnaire guide. This piece assumes you know the sequence and goes one level deeper.

What the analysis is trying to predict

Underneath every scoring rubric, a sponsor is estimating two numbers for your site: how many eligible patients you will actually enroll, and how much risk you add to the study. Everything on the questionnaire is a proxy for one of those. The sponsor's feasibility team — sometimes in-house, often at the CRO — compares sites side by side, and the protocol's enrollment target gets allocated across the sites that score best. A site that promises forty patients and a site that promises eight are not competing on the same axis; the sponsor needs both to be accurate, because the study's timeline is built on the sum.

The factors that carry the most weight

The exact weighting varies by sponsor, protocol, and therapeutic area, but the same handful of factors show up on nearly every questionnaire.

Eligible patient population. Not how many patients you see with the condition — how many would pass the protocol's inclusion and exclusion criteria, and how many of those you could realistically consent. Sponsors tend to discount optimistic numbers, and many apply their own haircut to whatever a site reports. A count backed by an EMR query with a stated date is worth far more than a round number from memory. How to build that count is covered in building a patient database for feasibility.

Enrollment track record. If you've run studies, the sponsor wants target versus actual enrollment for each, and whether you finished on time. A site that hit 80% of a realistic target reads better than one that promised double and delivered half. For a new site this factor is absent rather than negative — which is exactly why the population number has to carry more weight for you.

Investigator qualifications. The PI's specialty, experience in the indication, GCP training currency, and any audit or inspection history. Sub-investigators count. See how to become a Principal Investigator for what sponsors expect to find.

Staffing and bandwidth. Named coordinators, how many studies each carries, and whether the team has worked the kinds of procedures the protocol requires. Sponsors are wary of a site whose coordinator is already carrying a full load of studies, and equally wary of a site with no named coordinator at all.

Competing studies. Trials you're running that recruit from the same patient pool. Under-disclosing here is the fastest way to lose credibility if it surfaces later at the qualification visit.

Infrastructure. Space, equipment, drug storage, lab capability, and records. Mostly pass/fail — a protocol that needs –70°C storage either can run at your site or it can't — but gaps discovered late are expensive, so sponsors check early.

Start-up speed. How quickly you turned the CDA and questionnaire around, your IRB path, and your contracting history. Turnaround time is visible to the sponsor whether or not they track it formally, and it's the one factor entirely within your control on the day the questionnaire arrives.

How scoring usually works

Many sponsors and CROs score feasibility against some form of rubric — factors rated, ratings weighted, sites ranked — though the specifics are rarely shared with sites. Some use a numeric score; others sort sites into tiers (strong fit, possible, decline). The output feeds site selection along with practical constraints — geography, how many sites the budget allows, and whether the sponsor already has a relationship with you.

Two things follow. First, a mediocre answer in a heavily weighted category (population) outweighs a strong answer in a lightly weighted one (parking). Second, the sponsor is comparing you to other sites, not to an absolute bar — in a crowded indication the bar is higher; in a hard-to-recruit one, a modest but credible site can be selected over a flashier one that over-promised.

Run the analysis on yourself first

Before a questionnaire ever arrives, you can score your own site against the factors above. A practical self-assessment:

1. Population, by condition. Pull aggregate counts for your top five to ten conditions from your EMR, with the query date. For each, estimate what fraction typically meets common inclusion criteria (age bands, disease severity, no competing medications) — a rough funnel, honestly stated.
2. Investigator readiness. Current CV, license, GCP certificate and when it is due for renewal. If any of those is out of date, that's the first thing to fix.
3. Staffing. Who coordinates, how much capacity they have, and what procedures they've done. If the answer is "nobody yet," that's a plan to make, not a reason to stop.
4. Infrastructure inventory. Rooms, equipment, storage, lab, records system. Write it down once; it answers the same section on every questionnaire.
5. Speed. How fast could you return a CDA and a questionnaire today? If the honest answer is weeks, the bottleneck is usually legal review or who owns the task — both fixable.

The free Get Site-Ready checklist walks through this in order, with the outside pieces (training, registrations, IRB path) included.

Where sites lose the analysis

The common failure modes are consistent enough to list:

• Inflated patient counts. Probably the most damaging mistake, because the sponsor's whole plan depends on it and under-enrollment follows you to the next study.
• Vague staffing answers. "We'll hire a coordinator" reads as "we don't have one."
• Slow turnaround. A questionnaire returned in three weeks tells the sponsor how start-up will go.
• Undisclosed competing studies. Discovered at the qualification visit, it reopens every other answer.
• Expired credentials. A lapsed GCP certificate or license found by the sponsor can end the conversation on a technicality.

None of these is about the quality of your medicine. They're about whether the site looks organized — which is why organization is the cheapest competitive advantage in feasibility.

The analysis is yours to use too

Feasibility runs both ways. A protocol that would need 40 eligible patients from a pool where you have 30, or that would swamp your one coordinator, is a study to decline. Running the same analysis on the sponsor's protocol — against your real numbers — is how you avoid the under-enrollment that damages a site's reputation. Saying no to a poor fit is part of building the track record that wins the good ones.

Keeping those answers current is what feasibility software is for: profile, population, documents, and the questionnaire tracker (on the Site plan) in one place, so the analysis is already done when the sponsor asks.

---

This is general information, not legal or regulatory advice. Feasibility criteria and scoring vary by sponsor, CRO, and protocol.