Every month, you must be able to show which literature you searched, what came back, and why each record was decided the way it was.
That is a standing obligation, not a project. Under GVP Module VI a marketing-authorisation
holder must run a documented, repeatable literature search on a fixed schedule and keep the
screening record for inspection. We execute the search, screen every record that returns, and
hand you a signed record showing exactly what was searched and what was found.
human records, medicine implicated in a reported adverse event
Excluded — every one with a reason
37
60 − 23 = 37; no record dropped silently
Search window
36months
2023-11-01 → 2026-10-31, fixed and reproducible
Proof, not a promise
A real screening record, step by step
Below is an actual run over public literature: one medicine, a 36-month window, every retrieved
record screened and reasoned. Walk one record through the three stages it went through. The
search string, the raw API responses and the protocol are all published — read them, then
reproduce them.
Record journey
Loading the signed record…
Stage 1 — retrieved from the database
Excluded — X2
X2 — NO_NEW_CASES: a review, editorial, comment, guideline or meta-analysis carrying no new case data.
INCLUDE — INCLUDED: human record in which the medicine is implicated in a reported adverse event, tolerability or safety finding.
PMID
42888253
Safety vocabulary
adverse, event, events, suicidal
Implicating sentence
“Although previous studies have identified a significant signal for suicidal ideation associated with daridorexant when compared with all other drugs, the significance when compared specifically with insomnia medications and the potential differences in reporting patterns of suicidal ideation across subgroups remain unclear.”
The record above is signed with the Ed25519 attestation key published at
/v1/veritrack/public-key. The hash is fetched from
sample_record.sig.json when this page loads — it is not typed into
the page. Any change to the record breaks the hash.
60 retrieved − 23 flagged = 37 excluded
Exclusions in the published run, by reason code
Code
Reason
n
X2
NO_NEW_CASES: a review, editorial, comment, guideline or meta-analysis carrying no new case data.
20
X6
STUDY_PROTOCOL: a study protocol or design paper reporting no results; it can contain no case.
2
X3
NO_SAFETY_DATA: the record reports no adverse event, tolerability or safety outcome.
1
X4
DRUG_NOT_IMPLICATED: the medicine is mentioned, but no reported adverse event is attributed to it.
14
These counts are the real figures from the published run. The screening code aborts rather
than emitting a record whose counts do not reconcile.
The rule order is enforced, not asserted
Exclusion codes are applied in a fixed order — the first that matches is the recorded
reason:
X1 → X2 → X6 → X5 → X3 → X4 → INCLUDE
An automated check compares the published order against the order the screening code
actually applies, so the protocol you read is the protocol that ran.
The search strings, the databases, the date window and the inclusion and exclusion rules —
executed on a fixed schedule against PubMed/MEDLINE and the markets you state. The protocol
is the same document your auditor reads, and it is versioned.
screening_protocol.html
Every record screened, with its reason
Every retrieved record gets a decision and a reason. No silent drops: the count of
records screened always equals the count of records retrieved, and the record says so.
Anything that cannot be screened cleanly is routed to manual review, not dropped.
sample_record.json
Case candidates, flagged with citations
Suspected adverse-event and special-situation case candidates, each with its full citation,
a source link, and the exact sentence that implicated the medicine. The sentence is shown
because the sentence is what your causality assessment will turn on.
sample_record.html
An audit-ready monthly record
The protocol, the run, the counts, the decision for every record and the exclusion-reason
breakdown — assembled into a document built to survive an inspection question, and signed
with an Ed25519 attestation over its canonical form.
sample_record.sig.json
Retention that answers the inspection
The signed record, the search strings as executed, the raw API responses and the screening
decisions are retained with the record, so a third party can re-run the search and
reproduce the corpus. Your QPPV holds the file; we keep the receipt.
sample_record.json
Pricing
Two ways to buy the same work
Both tiers produce the same signed record. They differ in scope — how much of the
literature is yours to monitor rather than already handled by the EMA.
For a medicine whose literature is not on the EMA monitored list
Global screening
€690 per product per month
3 months prepaid — €2,070
For an active substance outside the EMA's monitored list, so its literature is not already
handled through EudraVigilance. Global database coverage, same protocol, same signature.
Stripe checkout, EUR, invoice generated automatically. The checkout collects the product
name (INN) and the markets, so the scope is declared by the buyer.
For local-language and regional journals in one named country
Local-language screening
€190 per product per country per month
3 months prepaid — €570
Per country. Portuguese, Spanish, Italian, Polish and other national journals screened
against a documented protocol, with the same signed record as the global tier.
Stripe checkout, EUR, invoice generated automatically. The checkout collects the product
name (INN) and the countries/markets, so the scope is declared by the buyer.
Multiple products are priced per product per month — three products on the global tier is
€2,070 per month. Payment is monthly in advance, prepaid three months.
What we do
Execute the documented search protocol on the fixed schedule.
Screen every retrieved record and record the reason for each decision.
Flag suspected adverse-event and special-situation case candidates with citations.
Produce the monthly screening record and the signed, hash-anchored log.
What we do not do
We do not make causality, expectedness or seriousness determinations.
We are not the QPPV. We do not sign as, or act as, your Qualified Person Responsible for Pharmacovigilance.
We do not submit regulatory reports on your behalf.
We do not replace the marketing-authorisation holder's own literature-search record.
The determination, the expectedness call and the signature are your qualified person's.
We supply the evidence, and we never touch the conclusion.
Start
Send one product and its markets
One product name (INN) and the markets it is authorised in. We reply with a one-page scope
and the sample screening record already published above. No form, no funnel — it reaches a
person.