Remote eConsent under ICH E6(R3): proving the investigator was involved
A participant signs an electronic consent form at their kitchen table. The platform records the signature, the timestamp and the form version. Every element of that record concerns the participant. Nothing in it concerns the investigator, whose involvement is the element a Good Clinical Practice inspector will ask about first.
How do you prove an investigator was involved in remote eConsent?
You mostly cannot. Decentralised consent records a participant signature, a timestamp and a consent version. It cannot show that a qualified investigator conducted the discussion, which is the substantive requirement — informed consent is a conversation, and the artefact records only its conclusion.
- eConsent platforms are architected around the participant's signature event. Investigator involvement is a separate administrative sign-off, frequently hours or days later.
- The measurable proxy is the interval between participant signature and investigator sign-off. It is exportable from every eConsent platform and nobody computes it.
- A dual-principal consent receipt binds participant and delegated investigator to the identical consent version hash within a defined discussion window.
Part of Pharma and GxP identity
The two records that never meet
Informed consent is not a document. It is a conversation, evidenced by a document. Regulations across jurisdictions are consistent on this point: the participant must have an opportunity to ask questions of a qualified person, and the consent must be obtained by the investigator or a properly delegated designee.
In a decentralised trial, those two things live in different systems. The participant's signature lives in the eConsent platform, bound to a form version and a timestamp. The delegation of authority lives in a delegation log, usually a spreadsheet or an eISF document. The discussion, if it happened, lives in a video call platform's attendance record or nowhere at all.
No artefact connects them. A sponsor asked to demonstrate investigator involvement for a specific participant assembles three exports and an argument.
Compute the interval
Before proposing a mechanism, establish the size of the problem in your own studies. Both timestamps exist.
| Interval | Reading | Inspection exposure |
|---|---|---|
| Same session, under 10 minutes | Contemporaneous involvement, plausibly the same call | Low |
| Under 24 hours | Same-day review; discussion timing unevidenced | Moderate |
| 1 to 7 days | Administrative sign-off decoupled from any discussion | High |
| Over 7 days | Batch clearing of a queue | High, and visible in the data |
There is no published benchmark for this and none is asserted here. The distribution across your own studies is the finding, and it tends to surprise clinical operations teams who assumed sites were signing contemporaneously.
What ICH E6(R3) changes about the framing
E6(R3) modernises GCP around risk proportionality and explicitly contemplates technology-enabled trial conduct. That is helpful — it removes the argument that remote consent is inherently suspect. It also raises the bar on evidencing that the quality factors critical to participant protection were actually managed, which puts the investigator-involvement question squarely in scope.
The practical consequence is that saying our SOP requires investigator involvement is a weaker answer in an E6(R3) world than it was under E6(R2), because the framework asks what evidence supports the assertion, not whether a procedure exists.
The Dual-Principal Consent Receipt
One canonical statement, signed by two principals, within a bounded window.
| Field | Content | What it establishes |
|---|---|---|
consent_version_hash | Hash of the exact consent document rendered to the participant | Both parties refer to the same version, resolving version-control disputes |
participant | Participant credential assertion | The participant signed |
investigator | Investigator or designee credential assertion | A named, delegated individual was involved |
delegation_ref | Reference to the signed delegation authorising that designee | Delegation is evidenced, not asserted in a spreadsheet |
window | Maximum interval between the two assertions | Contemporaneity becomes a constraint rather than a hope |
The design deliberately does not require the two signatures to be simultaneous. Consent discussions legitimately pause: a participant wants to consult family, or the call drops. A bounded window — typically the same calendar day — preserves flexibility while excluding the batch-clearing pattern that creates the exposure.
Delegation, done properly
The delegation log is the weakest artefact in most trials. It is a document listing who may do what, maintained by a coordinator, updated by hand, and frequently reconstructed at close-out.
Expressing delegation as a signed, scoped, time-bounded object fixes several problems at once: the principal investigator signs it, the scope names the specific activities, it expires, and it can be revoked when a sub-investigator leaves the site. A consent receipt referencing it inherits all of those properties.
The jurisdictional wrinkle
Requirements on contemporaneity differ. Some regulators expect the investigator to be present during the consent discussion; others accept documented review. This article does not flatten that difference, and any implementation should carry the window as a configurable parameter per study and per region rather than a single global value.
What is common across jurisdictions is the expectation that the sponsor can demonstrate what happened. A receipt satisfies that expectation regardless of where the window is set.
What it protects
The downside risk in consent findings is not a fine. It is exclusion of participant data from the analysis, which in a pivotal trial converts a documentation issue into a statistical one. Sites with systematic consent findings face further consequences.
Against that, the cost is one additional device gesture per consent from a person who is already required to be involved. That is an unusually favourable ratio, which is why this is worth doing before an inspection rather than after one.
Why remote makes this harder
| Element | In person | Remote |
|---|---|---|
| Investigator presence | Observable | Asserted |
| Participant identity | Visual | Document or credential |
| Discussion occurred | Assumed from the setting | No artefact |
| Version signed | Physical document | Good — versioned and logged |
Decentralised consent is genuinely better on version control and worse on presence. Recognising that lets a sponsor strengthen the weak half rather than defending the whole approach.
Objections and honest limits
“The platform records the session.” Some do, and a recording is evidence a session occurred rather than that a qualified investigator conducted it. It also raises its own retention and privacy questions.
“Investigator oversight is delegated to site staff.” Delegation is permitted within limits, and the record should say who actually conducted the discussion — which is precisely what is missing.
Strengthening remote consent
- Capture two signatures, not one. Participant and the person who conducted the discussion.
- Record the conductor's qualification. Linked, not assumed.
- Bind the consent version to both. So a later version cannot inherit an earlier consent.
- Keep it verifiable outside the platform. Trials outlive platform contracts.
Terms used here
- eConsent
- Electronic informed consent, including remote and decentralised administration.
- Investigator
- The individual responsible for the conduct of a trial at a site, including the consent discussion.
- Decentralised trial
- A trial conducted wholly or partly outside traditional sites, with consent obtained remotely.
Frequently asked questions
Does this require the investigator to be on the call? No. It requires them to sign the same consent version within a window you define, which may be configured to match the expectations of the regions in your study.
What about participants who cannot use a device credential? Provide an assisted path at the site and record it as an exception. A consent mechanism that excludes participants is not an improvement, and the exception record is itself useful evidence.
Does the receipt contain personal data? It carries a hash of the consent document version and credential assertions. It does not carry participant health information.
Can this replace the eConsent platform? No. It runs alongside it, adding a bound investigator assertion to the platform's existing record.
What does eConsent do better than paper? Version control. It reliably records which consent version was signed, which paper handles poorly.
What does it do worse? Presence. Whether a qualified investigator conducted the discussion is asserted rather than recorded.
What is the minimal fix? A second signature from the person who conducted the discussion, linked to their qualification.
Where this fits in Manav
Manav binds the signer to the exact record being certified, on a credential under their sole control, and produces a receipt an inspector or a partner can verify without access to the originating system.