Pillar 3 of 6 · Cohort
Can you actually find the patients?
How many eligible patients exist, how much your own eligibility criteria shrink that pool, and how wide the honest range around the answer really is.
Can you find the patients?
What it answers
Can you find the patients?
Recruitment is where feasibility studies are usually wrong, and usually optimistic. This pillar walks the funnel from a population denominator down to patients who could actually be enrolled at a site — and shows you the range, not a single confident number.
How many patients exist?
An epidemiology-anchored denominator for the indication, in the countries you are considering, with the snapshot it came from and the uncertainty it carries.
How many survive your criteria?
What your inclusion and exclusion list does to that pool — scored against the criteria of comparable trials using the identical classifier, so the comparison is like for like.
How many can you actually enrol?
The step from eligible to enrollable: screen-fail priors, registry recruitment signals, and where a licensed real-world data source can give you a claims- or EHR-derived count instead of an estimate.
Inside the pillar
3 analyses, each one named and cited
These are the analyses that actually run — not a category list. Each states what it computes and what it needs, because an analysis that quietly degrades is worse than one that reports it could not run.
Cohort sizing
Walks an epidemiology-anchored funnel from population denominators down to an enrollable cohort, applying a screen-fail prior and registry recruitment and participant-flow signals, then runs a Monte Carlo over the funnel so the output is a distribution with its drivers rather than a point estimate. Every step states its own basis and its own warnings.
Needs
An indication and, when you have them, a phase and a target enrolment. When the study does not state them the analysis records them as absent rather than substituting a default, and declines to compute a site count on made-up inputs.
Eligibility complexity
Scores your inclusion and exclusion criteria on an eight-axis lexicon, and scores a real peer set of same-phase interventional trials with the identical classifier over the identical splitter — so your criteria land as a percentile against comparable protocols rather than as an uninterpretable absolute count.
Needs
Your eligibility criteria as text, and the phase (phase 2 or 3 is the fallback peer filter, and the filter used is recorded on the output).
Real-world cohort
Queries licensed real-world data vendors for a claims- or EHR-derived count of patients who would meet your criteria, so the eligible pool is observed rather than modelled.
Needs
A connected real-world data vendor. Without one, the analysis returns an explicit access-required state; it never silently falls back to the public surrogate and presents it as real-world evidence, and a hard failure on a connected vendor is reported as a failure rather than as an upsell.
Where the evidence comes from
Public sources, named
The epidemiology layer is a bundled snapshot and is cited as one — never as a live feed from its steward. Where a prior entered the arithmetic, the assessment lists that prior; where a fetch failed, the failure becomes a visible warning rather than a silent zero.
PubMed
NCBI / U.S. National Library of Medicine · Daily
Biomedical literature index of 36M+ citations — used to ground indication baselines and standard-of-care evidence.
pubmed.ncbi.nlm.nih.gov (opens in a new tab)ClinicalTrials.gov
U.S. National Library of Medicine · Daily
Registry of public and private clinical studies worldwide — enrollment, status, sponsors, sites, and outcomes.
clinicaltrials.gov (opens in a new tab)AACT
Clinical Trials Transformation Initiative (CTTI) · Daily
Relational, queryable database of all ClinicalTrials.gov content — the analytical backbone for site and enrollment statistics.
aact.ctti-clinicaltrials.org (opens in a new tab)Also read by this pillar
What you get
What lands in the assessment
Every funnel step names its basis — the prior, the snapshot, or the registry query behind it — and every external fetch that failed appears as a warning on the tile.
- A Cohort tab with the sizing funnel, its Monte Carlo range, and the basis stated for every step
- An eligibility-complexity percentile against a named peer set, with the axes that drove it
- A real-world count where a vendor is connected, or an explicit access-required state where one is not
Read by these expert roles
A role drafts from this pillar; a human reviews and signs off. The sign-off is a 21 CFR Part 11 e-signature on a hash-chained record.
How the roles workHonest limits
What this pillar cannot tell you
Every assessment method has a boundary. Publishing ours is the point — a number you cannot check is worth less than a gap you can see.
A funnel is not a recruitment guarantee.
Population denominators are national, snapshot-dated and coarse. A rare subtype, a biomarker-defined population, or a competitive indication can leave you with a fraction of what the funnel implies.
Screen-fail priors are borrowed.
They come from comparable trials, not from your protocol. If your criteria are unusual in a way the lexicon does not capture, the prior will be wrong in the direction of optimism.
The observed count is licence-gated.
A claims- or EHR-derived cohort requires a real-world data licence that you hold, not one we hold on your behalf. Without it you get a modelled surrogate, clearly labelled as one.
It does not know your sites.
Eligible patients in a country are not patients at your investigator sites. The step from national pool to site-level accrual is the sites and cost pillar, and it carries its own uncertainty.
The rest of the assessment
No pillar decides alone
Each pillar answers one question. The Verdict composes all six into a single go/no-go, with the risks that could still change it.
See a verdict you can actually check.
Send us a protocol — or just a molecule and an indication. We'll return a fully cited feasibility assessment you can trace, line by line, back to public data — yours to defend in a bid, take to your board or investment committee, or hand to a regulator.