Glossary

The vocabulary, defined without hand-waving.

Every term the assessment uses, in plain language — including the ones the industry uses loosely. Deep-link any definition.

Terms you can send someone

Glossary

The verdict

The terms the go/no-go call is made in.

PoS(Probability of Success)#: The historical share of comparable trials that cleared this phase — a base rate drawn from similar programmes, not a promise about yours. It is presented as directional evidence with its assumptions and comparator set visible, and is not a calibrated forecast.See also: Calibration, Evidence band
Go / No-Go#: The recommendation the assessment ends on, together with the drivers behind it. It is a synthesis of the five pillars beneath it, not an independent opinion — which is why every part of it can be traced back to a specific analysis and a specific public record.
Calibration#: Whether a confidence number has been checked against real outcomes. Uncalibrated means the score orders results — more evidence gives a higher score — but cannot be read as a percentage chance. Where a score is uncalibrated, the product says so rather than letting the percent sign imply otherwise.See also: PoS
Operational risk#: The delivery-side risks that sink trials which were scientifically sound: recruitment competition, site activation, amendment exposure, and sponsor-level signals such as disclosed delays in public filings.

Glossary

Patients & eligibility

How many patients exist, and how many of them your protocol would actually accept.

Prevalent population#: The number of people living with the condition in a country — a stock of cases, derived from a prevalence rate and a population denominator. It is not the number diagnosed, treated, or reachable, and the bundled epidemiology is illustrative point estimates rather than a live feed.See also: Eligible pool
Eligible pool#: Prevalent patients who pass every modelled eligibility criterion and the screen-failure prior. It is eligible, not recruitable: no reach, catchment or consent stage is modelled, so treat it as an upper bound on who could qualify.See also: Prevalent population, Screen failure, Eligibility complexity
Eligibility complexity#: How restrictive your inclusion and exclusion criteria are, measured against comparable trials in the same indication rather than in the abstract — for example, stricter than 82% of 173 peer trials. It is the analysis that turns "our criteria are quite tight" into a number you can argue about.See also: Eligible pool, Protocol amendment
Screen failure#: The share of screened patients who do not go on to randomise. Resolved from registry records for the indication and phase where they support it, and otherwise a stated prior — the basis label says which.See also: Basis
Real-world cohort#: A cross-check on the modelled cohort using published real-world evidence for the same condition, so the sizing does not rest on a single epidemiological chain.

Glossary

Sites, countries & cost

Where the trial runs, who runs it, and what it costs to do so.

Site density#: How many qualified investigator sites exist for this indication in a given geography — and how many other active trials are competing for the same patients at them. High density with high competition is a slower recruitment story than the raw site count suggests.See also: Competition
Competition#: The number of other trials with an active registry status recruiting the same condition in your enrolment window. More competitors means harder, slower recruitment at the same sites.See also: Site density
Priority tier#: The model’s ranking of a country or site for this trial. P1 means run here first, P2 is a strong backup, P3 only if needed, and SKIP means de-prioritise.
Cost band#: A low–median–high estimate built from public per-patient priors. It is a triangular band, not true percentiles, and public-data budgets should be treated as roughly ±50%. Useful for comparing designs and geographies; not a substitute for a CRO bid.
Operational benchmarks#: What comparable trials actually achieved — activation timelines, enrolment velocity, and duration — drawn from registry records rather than from a vendor’s internal averages.

Glossary

The commercial case

What the asset is worth if the trial works.

eNPV(expected Net Present Value)#: The value of the programme after discounting future cash flows to today and weighting them by the probability of getting there. It is the number that makes a clinical decision a financial one — and because it multiplies through a probability of success, it inherits every caveat that number carries.See also: PoS, TAM / SAM / SOM
TAM / SAM / SOM(Total, Serviceable and Obtainable Addressable Market)#: Three nested market sizes. TAM is everyone with the condition; SAM is the share your label, geography and route could serve; SOM is the share you could realistically win against the standard of care and the competition. Quoting TAM alone is the oldest way to make a market look bigger than it is, which is why the assessment reports all three.See also: Standard of care
Value-based price#: A price anchored to the clinical benefit over the current standard of care, rather than to a competitor’s list price. It sets the ceiling the commercial case has to work under.See also: Standard of care
Standard of care#: What patients receive today. It is the comparator your trial has to beat clinically and your price has to be justified against commercially — so it constrains both the design and the market at once.
Patent landscape & exclusivity#: The patent estate around the asset and the runway of market exclusivity remaining. A strong clinical case on a short runway is a different investment from the same case with a decade of protection.

Glossary

Evidence & provenance

How the product tells you how much to trust a number.

Evidence band#: How much independent evidence backs a number, from source diversity and record counts: no data → weak → single source → converging → triangulated. It is a statement about the evidence, not about the outcome.See also: Basis, Citation
Basis#: Where a modelled input came from: stated (you entered it), published (a cited figure), computed from registry records, curated (a hand-picked constant with a citation), heuristic (one without), an assumption, or not modelled at all. It is how you tell a fact from a knob.See also: Evidence band
Citation#: The provenance attached to a figure: the public source, the query that produced it, the date the record was read, and a link back to the record itself. Enough to reproduce the number without asking us.See also: Snapshot
Snapshot#: The pinned version of a public record at the moment the run read it. Public data moves; a snapshot is what lets a figure be defended months later against the record as it stood rather than as it is now.See also: Citation
Evidence ledger#: The per-study list of every connector the run touched and what it returned — including the ones that returned nothing. A sparse assessment is visible as sparse rather than presented with the same confidence as a well-evidenced one.

Glossary

Governance & compliance

The record-keeping vocabulary a regulated decision runs on.

21 CFR Part 11#: The US FDA regulation governing electronic records and electronic signatures in regulated work. In practice it is what makes an electronic approval count: records must be attributable, tamper-evident and retained, and signatures must be bound to the record they signed.See also: Electronic signature (§11.70), Hash-chained audit trail
Electronic signature (§11.70)#: A signature bound to the record it approves, so it cannot be excised, copied or transferred to another document. In the product, signing captures a mandatory reason, pins the citation versions the reviewer saw, records a content hash of the signed artefact, and writes a signing event to the audit chain.See also: 21 CFR Part 11, Hash-chained audit trail
Hash-chained audit trail#: An append-only record in which each entry references the cryptographic hash of the entry before it. Because every entry depends on its predecessor, a retroactive edit anywhere breaks the chain and is detectable — which is what makes the record tamper-evident rather than merely stored.See also: 21 CFR Part 11
Protocol amendment#: A formal change to a trial protocol after it has been approved. Amendments are expensive: they can require re-approval, re-consent and re-training across every site, and substantial ones are among the most reliable predictors of delay. The Design pillar estimates amendment risk from what comparable trials had to change.See also: Eligibility complexity
Re-signature#: When a cited source behind a signed deliverable is superseded, the task re-opens as requiring a fresh signature. The original signature stays in the record, attached to the evidence as it stood — so you can always tell whether a signed decision still rests on current evidence.See also: Electronic signature (§11.70)
Tenant isolation#: Each customer’s studies, documents, citations and audit chain are held separately from every other customer’s. Uploaded documents are retrievable only within the workspace they were uploaded to.

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.

Book a demo Sign in