When Should Informed Consent Be Obtained?

Most informed-consent guidance focuses on what a valid consent process must disclose. A separate, narrower question trips up study teams just as often: when does consent have to be obtained, and what happens when the timing isn’t a clean one-time event — because the protocol changes mid-study, or because the research population can’t consent at the moment intervention is needed? This guide covers three timing-specific rules under the US Common Rule: the prospective-consent requirement, re-consent triggers when a protocol is materially amended, and the narrow regulatory exception that allows emergency research to proceed without prior consent at all.

For what a valid consent process has to contain, see CASRAI’s companion guide Informed Consent in Research: What It Requires and How It Works. For the document itself, see ICF (Informed Consent Form) and Informed Consent Form: A Worked Example With Annotations. This page is specifically about timing — before, during, and in the emergency-research edge case.

The general rule: consent before any study-related procedure begins

The Common Rule states the timing requirement directly in the opening clause of 45 CFR 46.116: except where an exception or waiver applies, an investigator may not involve a human being as a subject in covered research unless the investigator has obtained the legally effective informed consent of the subject or the subject’s legally authorized representative (LAR). “Involve as a subject” is read broadly — it covers any study-specific intervention or interaction carried out for research purposes, not only the primary endpoint procedure. In practice this means:

Consent (or a properly IRB-approved waiver of it) must be in place before any research-only procedure is performed — a study-specific blood draw, an investigational drug dose, a research questionnaire, randomization, or any data collection that would not have occurred as part of routine clinical care.
Consent must be obtained with enough time and privacy for genuine deliberation — 45 CFR 46.116(a)(1) requires investigators to seek consent only under circumstances providing the prospective participant “sufficient opportunity to discuss and consider whether or not to participate,” which is why same-day, immediately-before-the-procedure consent is disfavored where it’s avoidable.
Routine clinical activities that would have happened regardless of the study — standard-of-care labs, a clinically indicated scan — don’t themselves require research consent, but once a study team uses that data or specimen for research purposes, or adds any study-specific procedure to a clinical visit, the consent requirement attaches to the research component.

Screening is a common edge case. Where determining eligibility itself requires a procedure or the collection of otherwise-private information beyond what’s needed for routine care (a research-specific screening lab, a detailed eligibility questionnaire), many IRBs require a short, separate screening consent before that screening occurs, distinct from the full study consent obtained later if the participant is eligible and agrees to enroll. Whether a separate screening consent is required, or eligibility can be assessed from information already properly available to the investigator, is an IRB determination specific to the protocol — there’s no single federal rule dictating the answer either way, which is why protocols spell out their screening consent approach explicitly for IRB review.

Re-consent: when a protocol amendment requires going back to participants

Consent obtained at enrollment doesn’t cover changes made to the study after that point. Two related but distinct regulatory mechanisms govern what happens when a protocol is materially amended mid-study:

IRB approval of the change itself

Under 45 CFR 46.108, an IRB must review and approve any proposed change to previously approved research — including a revised consent document — before that change is implemented, except where the change is necessary to eliminate an apparent immediate hazard to participants. This is the procedural gate: a study team cannot simply start using a revised consent form or protocol because a change seems minor. See CASRAI’s Protocol Amendment entry for how amendments are submitted and tracked, and Protocol Deviation for what happens when a change is implemented without that prior approval.

Telling already-enrolled participants

Separately, 45 CFR 46.116(b)(5) (and the parallel additional-elements provision in the 2018-revised text) requires that, where appropriate, the consent process disclose that participants will be informed of significant new findings developed during the research that may relate to their willingness to continue participating. This is the regulatory basis for re-consent: if an amendment introduces a new risk, meaningfully increases the magnitude of a known risk, reduces expected benefit, or otherwise changes information a reasonable participant would want in deciding whether to continue, already-enrolled participants generally need to be given that information and, depending on the IRB’s determination, asked to re-consent — sign a revised or addendum consent form — rather than simply continuing under the original signed form.

Not every amendment triggers re-consent. The IRB decides the appropriate mechanism based on the significance of the change:

Full re-consent (participant reviews and signs a revised consent document, typically at their next study visit) — for changes that materially affect risk, benefit, or the procedures a participant already agreed to.
Notification only (a letter or verbal update, without a new signature) — for changes the IRB determines are informational but not significant enough to affect a reasonable participant’s willingness to continue.
No participant-facing action — for administrative amendments (e.g., a new sub-investigator, a clarified procedure with no risk/benefit implication) that don’t change what a participant was told or agreed to.

Which category an amendment falls into is an IRB determination made at the time the amendment is reviewed under 46.108 — it is not something a study team decides unilaterally, and continuing to follow the original consent after a material change without IRB-directed re-consent is itself a protocol deviation.

The narrow exception: research without prior consent in emergency settings

A small, tightly bounded regulatory pathway allows specific emergency research to proceed without obtaining prior informed consent at all — not as a general emergency-research carve-out, but as a narrowly conditioned exception built for a specific scenario: testing an intervention for a life-threatening condition where, because of that condition, the individual cannot consent and no legally authorized representative can be reached in the available treatment window.

For FDA-regulated investigations of a drug, biologic, or device, this is governed by 21 CFR 50.24, “Exception from informed consent requirements for emergency research” — often referred to as EFIC. An IRB may approve this pathway only where all of the following are true: the subjects are in a life-threatening situation for which available treatments are unproven or unsatisfactory; obtaining consent is not feasible because the subject cannot consent as a result of the medical condition and the intervention must be administered before consent from an LAR is feasible; participation holds out the prospect of direct benefit; the research could not practicably be carried out without the exception; and, critically, the IRB has confirmed that the investigator has completed — before the study begins — community consultation and public disclosure to the community from which subjects will be drawn, plus additional safeguards including an independent data monitoring committee and a mechanism for individuals to decline participation in advance.

The parallel authority on the HHS/Common Rule side sits at 45 CFR 46.101(i): the HHS Secretary may waive the general consent requirements of 46.116 and 46.408 for research carried out on subjects who need emergency therapy and for whom, because of their medical condition and the unavailability of an LAR, no legally effective informed consent can be obtained. This waiver authority does not extend to research involving pregnant women, human fetuses, neonates, or prisoners.

Three things make this exception narrower than it’s sometimes assumed to be:

It is prospective and protocol-specific, not retroactive. Community consultation and IRB approval of the EFIC pathway happen before the study enrolls anyone under this provision — it cannot be invoked after the fact to justify treating a subject without consent on an ad hoc basis.
Consent is still sought as soon as it becomes feasible. Once a subject regains capacity, or an LAR becomes available, the investigator must seek consent for continued participation as soon as feasible; the exception covers only the window during which neither the subject nor an LAR could be reached in time to authorize the intervention.
It does not apply broadly to “emergency research.” Plenty of research involving emergency-department patients or critically ill participants uses ordinary consent (from the patient, once stabilized, or from an LAR) — EFIC/46.101(i) is reserved specifically for the subset of studies where the intervention window is too short for any consent pathway to work.

Frequently asked questions

Can a researcher collect any study data before consent is signed?

Not for research-specific procedures or data collection. Information already properly available for other reasons (e.g., existing medical records under an approved secondary-use protocol) is a separate question governed by its own consent or waiver terms. Where eligibility screening itself requires new research-specific information or a procedure, the IRB typically requires a separate, shorter screening consent before that screening occurs.

Does every protocol amendment require getting participants to sign a new consent form?

No. The IRB determines, when it reviews the amendment under 45 CFR 46.108, whether the change is significant enough to require full re-consent, informational notification only, or no participant-facing action at all. The trigger for re-consent is whether the change affects information a reasonable participant would want in deciding whether to continue — per 45 CFR 46.116(b)(5) — not whether any change was made at all.

Is the emergency-research exception the same as a waiver of consent for minimal-risk research?

No, and conflating them is a common mistake. The minimal-risk waiver/alteration of consent under 45 CFR 46.116(f), covered in CASRAI’s Informed Consent in Research guide, applies where research involves no more than minimal risk and the waiver won’t adversely affect participants’ rights. The emergency-research exception (21 CFR 50.24 / 45 CFR 46.101(i)) applies to a specific, higher-risk, life-threatening-condition scenario and carries its own separate, more demanding set of conditions, including mandatory community consultation before the study begins.

Who decides whether an amendment is “material” enough to trigger re-consent?

The IRB, not the study team. Amendments are submitted for IRB review under 45 CFR 46.108, and the IRB’s approval of the amendment includes its determination of what, if anything, must be communicated to or re-consented with already-enrolled participants.

When Should Informed Consent Be Obtained? Timing, Re-Consent, and the Emergency Exception