Prove your trial documents were fixed before the results were known
The Pre-Specification Certificate records exactly which protocol, analysis plan and other key documents a clinical trial had, and when, so anyone can later confirm they were not changed after the fact. Nobody has to take Graticus's word for it: the proof checks out independently.
The problem it solves
A trial's conclusions are only as trustworthy as the plan written before the data came in. If the primary endpoint, the analysis or the randomisation list could have been adjusted after anyone saw the results, the findings are open to doubt. Reviewers, regulators, journals and partners increasingly ask sponsors to show, not just say, that the plan came first.
Dates typed into a document, a file's "last modified" time or an email trail can all be changed or disputed. A Pre-Specification Certificate gives you evidence that cannot be quietly back-dated, including by us.
How it works
- Record the trial's milestones, such as first patient enrolled, database lock and unblinding. Planned dates are fine and can be corrected; every correction stays on the record.
- Add each document before its milestone. Your computer calculates the document's fingerprint, a 64-character code unique to that exact file. Only the fingerprint is sent. The document itself never leaves your computer. The fingerprint is then fixed in time in two independent ways: a signed timestamp from an independent timestamp authority, and an entry in the public Bitcoin blockchain, which nobody can alter afterwards.
- Issue a certificate and share it. The certificate lists every document, its fingerprint, when it was fixed, and whether that was before its milestone. Send it with the documents to whoever needs to check.
- Anyone checks it on the verification page, with no account, by dropping in the certificate and the documents. Each document shows green if it is exactly the file that was fixed, and red if even one character has changed. The check runs in their own browser, and the page also works saved to their computer with no connection to Graticus at all.
What you can use it for
Protocol before first patient
Show the protocol and primary endpoint were set before enrolment began, and that any later version came with a recorded reason.
Analysis plan before unblinding
Show the statistical analysis plan was fixed before anyone saw unblinded data, the point where "outcome switching" concerns usually arise.
Randomisation list before randomising
Show the allocation list existed, unchanged, before the first patient was randomised.
A clean amendment trail
Every new version is recorded alongside the one it replaces, with its reason and the date it takes effect. Nothing is overwritten or deleted.
Database lock and topline results
Fix the database-lock certificate and the topline results at the moment they exist, so the sequence of events is beyond dispute.
Partner attestation
Invite the CRO or statistician who wrote a document to confirm, on the record, that it is theirs in this exact form.
Who it is for
| Who | How it helps |
|---|---|
| Sponsors, including small biotech and pharma | Evidence of good practice to hand to partners, investors, regulators and journals, without sharing confidential documents with us. |
| CROs, statisticians and trial sites | Invited onto a trial record to add the documents they prepared and attest the ones they wrote. |
| Journal editors and peer reviewers | Confirm that the analysis reported matches a plan fixed in advance. |
| Regulators and auditors | Check document integrity and sequence independently, using open standards. |
| Investors, licensing partners and acquirers | Due diligence on a programme's data: was the plan fixed before the readout? |
What a certificate proves, and what it does not
- Each document existed in exactly this form no later than the time shown.
- Whether that time was before the milestone it is listed against.
- Any version that replaced it was recorded, when, and with what reason.
- That the milestone dates are true. They are the sponsor's own statement, and the certificate labels them that way.
- Anything about the trial's results, the quality of its analysis, or whether the plan was followed.
- Who wrote each document, except where a named partner has attested it.
Questions
Do you see or store our documents?
No. The fingerprint is calculated in your browser, and only the fingerprint is sent. A fingerprint cannot be turned back into the document. To check a certificate, a verifier needs the documents from you.
What if Graticus disappears?
The certificates still check out. The verification page can be saved and opened from any computer, and the proofs rest on a public timestamp authority and the Bitcoin blockchain, not on our servers. The page also explains how to check every proof with standard tools.
Why does a re-saved copy show red?
The check compares the exact file. Opening a document and saving it again, even unchanged, alters its contents, so its fingerprint no longer matches. Keep the copy you fixed in a folder of its own and send that copy.
Which timestamp authority do you use?
Currently FreeTSA, an independent authority whose signed timestamps anyone can check. It is not an EU eIDAS-qualified provider. Each certificate also carries a Bitcoin proof, which does not depend on any authority.
How do I start?
Sponsors sign in through the Graticus Platform. Trial partners join by invitation from a sponsor. To ask about access, write to hello@graticus.com.