How an ICF is generated from an M11 protocol
Which protocol content feeds which consent sections, how reading level and translations are handled, and what happens on an amendment.
An informed consent form repeats a large part of the protocol in plain words. In most teams it is a separate file, written after the protocol and updated by hand each time the protocol changes. When the protocol is held as structured fields, the consent form can be drafted from those fields instead. This article describes how that works in Trelice, and where people still do the work.
Why can a consent form be generated from a protocol?
ICH M11 was designed with re-use in mind. One of the template's design principles reads: "The protocol is a rich source of information that can be re-used as a part of the clinical trial management and review process." The template puts the same information in the same place in every protocol, and the guideline names trial participants among the people a protocol template should serve.
That matters for consent because a generator has to know where the procedures, the visit schedule, and the risks are. In a free-text protocol, a person has to find them. In a structured protocol, they are fields that software can read. Trelice generates its documents from those fields, not from copied text.
Which protocol content feeds which consent sections?
In Trelice, the master consent form has 16 sections. It opens with a title page and a key information summary and closes with the signature block. Between them are an introduction, study procedures, duration, risks and discomforts, benefits, alternatives, confidentiality, data protection, compensation and costs, insurance, contact information, voluntary participation, and optional studies.
When the form is created, Trelice copies matching information from the protocol into those sections. That includes the study title, the procedures, the visit schedule, and the risks. Risk language also draws on the Investigator's Brochure, where the product's identified and potential risks are kept.
Each section is a short form with labeled boxes, the same idea as a structured protocol section. Sections marked as required must have content before the form is complete. The writer reviews every section and completes what the protocol could not fill.
You create one master per region: United States, European Union, or global. The region decides which regulatory sections the form includes.
How is reading level handled?
A consent form has to be easy to read. Trelice measures each drafted section on the Flesch-Kincaid scale and shows the result as a badge in the editor. The target is a grade level of 3.0 or below.
The check is a gate, not a suggestion. When a drafted section is above the target, it cannot be submitted for review until it is rewritten or regenerated. Writers shorten the sentences and swap medical words for everyday ones.
Once a section passes, it moves through the same review workflow as a protocol section. The writer assigns a reviewer and submits. Editing is locked during review, and the reviewer approves or requests changes.
How do translations and country versions work?
Translations are added per section. The writer picks a language, then types the translation or, if the organization has a translation service set up, asks Trelice to draft it. A translation does not count until a qualified person has reviewed it. Only a colleague with the Language Specialist role and the matching language specialty can do that review.
A country consent form is a full copy of the master, adapted and translated for one country and language. When you choose a country, Trelice lists the languages that country requires and shows local regulatory notes, such as whether back-translation is required. Trelice flags that requirement. The back-translation itself is arranged by your team.
Country forms go through the same section-by-section review as the master. Who translated, who reviewed, who approved, and when are all recorded in the audit trail.
What happens to the consent form on a protocol amendment?
An amendment in Trelice is made once, in the protocol, with tracked changes and a written reason for each change. The amendment list carries re-consent flags that mark the amendments whose consent forms must follow.
A country version is a snapshot of the master at the moment it was generated. If the master is edited later, each affected country version shows a warning that the master has been updated and its content may be outdated. The team then compares the two, carries the changes over, or regenerates the country version once the master is final again. Consent forms are also flagged as stale when a new edition of the Investigator's Brochure changes the risk information they draw on.
It helps to be exact about what is automatic. The warnings are automatic. The decisions are not. Whether to update a form, send it back through review, and re-consent participants stays with the study team.
What if the protocol was not written in Trelice?
Some protocols exist only as a PDF from a partner or an older system. The ICF Generator takes that PDF, extracts its text, and drafts a consent form for the region you choose. This step uses AI to read the extracted text and draft each section. The guide for it is direct: an AI draft is a starting point, never a finished document, and every section should be read against the source protocol before it goes to review.
The result opens in the same editor as any other consent form, with the same reading-level check, the same section review, and the same export to PDF and Word.
What should you check before relying on a generated consent form?
Generation removes retyping. It does not remove responsibility. The form still goes through your own review and your ethics submission, so three checks are worth building into the process.
First, read every section against the approved version of the protocol, not a draft. Second, confirm that each translation was reviewed by a specialist qualified in that language. Third, after any amendment, open each country version and look for the outdated warning before a form goes to a site.
Related on this site
- Downstream documents
- Guide: the master consent form
- Guide: translations and country versions
- Guide: the ICF Generator
Sources
- EMA: ICH M11 Guideline on clinical electronic structured harmonised protocol (CeSHarP), Step 5 (PDF)
- EMA: ICH M11 guideline, clinical study protocol template and technical specifications