Overseeing a Consent Process You Did Not Run

Part of Principal Investigator (PI) Training: Consent as the Investigator's Duty.

About 4 minutes. Free, in your language, no account needed.

Why this matters

Consent findings are among the most common at inspection, and the great majority are documentation and version-control failures rather than ethical ones. The discussion was fine. The record, the form version or the re-consent was not.

Those are supervision failures, and they are cheap to prevent — if somebody is looking.

The version-control problem

Section 2.8.2 requires the IRB or IEC to give written approval of the consent form and other written information before use. Every amendment produces a new version, and every new version creates a window in which the wrong document is sitting in a folder at your site.

Three controls close it:

1. One controlled location. A single current version of the consent form, in one place, physical or electronic. Superseded versions are removed the day the new one is approved — not archived alongside it "just in case".
2. A version log. Version number, approval date, date first used at your site, date withdrawn. One page, and it answers the inspector's question in ten seconds.
3. A named owner. One person is responsible for the swap. Distributed responsibility means the swap happens twice or not at all.

Then check it. Pick three recent consents and confirm the version used was current on that date. If you do this quarterly you will find your own problems.

Re-consent when something changes

Sections 2.8.11 and 2.8.12 cover new information: where information becomes available that may be relevant to a participant's willingness to continue, they are informed in a timely manner, revised materials receive IRB/IEC approval, and the participant's willingness to continue is documented.

This is the obligation that most often slips, because it lands mid-trial when everyone is busy and the trigger comes from outside the site — a safety update, a Brochure revision, an amendment.

Make the trigger mechanical. When any of these arrives, one person asks a single question in writing: does this require re-consent of currently enrolled participants, and by when? Log the answer either way. A documented decision that re-consent was not required is a defensible position; silence is not.

And re-consent is a discussion, not a signature run. A participant being told about a new safety finding is being given the opportunity to reconsider. That is what the process is for.

Situations that need you in the room

Most consents can be conducted by a trained delegate. Some raise the risk enough that your involvement is the proportionate supervision §2.3.2 describes.

• Impaired capacity. Consent by a legally acceptable representative under §2.8.9, with the participant informed to the extent of their understanding and assent obtained where feasible.
• Minors. Age-appropriate information and assent, with re-consent when a participant reaches the age of legal consent during the trial.
• Participants who cannot read. An impartial witness under §2.8.6, present for the discussion and signing to confirm the information was accurately explained and consent freely given.
• Emergency enrollment, where the protocol provides for it, with consent to continue obtained as soon as possible (§2.8.13).
• Anyone the Glossary would treat as vulnerable, where the capacity to give consent freely may be unduly influenced by expectation of benefit or fear of retaliation.

Decide at start-up which of these your trial will encounter, and write down who does what. Deciding in the moment, with a distressed family present, is how sites get it wrong.

A ten-minute monthly review

The whole of consent oversight fits into one recurring habit. Once a month, take every consent since the last review and check five things:

• The form version was current on the date used.
• Consent preceded every trial-specific procedure.
• Participant signature and date are in the participant's own hand.
• The consent note describes a discussion, not just a signing.
• Any required re-consent has happened, or a documented decision says it was not required.

Five checks, a few minutes per participant, and it catches nearly everything that inspections find.

Key takeaways

• Most consent findings are version control and documentation, not ethics — and both are supervision problems.
• Run one controlled current version, a version log and a named owner; verify quarterly against real consents.
• §2.8.11–§2.8.12: make new information a mechanical re-consent trigger, and log the decision even when the answer is no.
• Be personally involved where capacity, minors, illiteracy, emergency enrollment or vulnerability are in play, and decide the approach at start-up.