Evidence explainer

Diabetes and metabolic health

What Informed Consent Really Means

Informed consent in a clinical trial is a conversation, not a signature. The form records part of it; understanding, time, questions, and freedom from pressure decide whether it means anything.

Fully reviewed by Jasaman (Jasmin) Tojjar, MD, PhD

On this page
  1. Consent begins before the form
  2. The core information
  3. Therapeutic misconception
  4. Capacity is decision-specific
  5. Children and assent
  6. Voluntariness and undue influence
  7. Understandable and accessible communication
  8. Electronic and remote consent
  9. Consent is ongoing
  10. Withdrawal and already collected data
  11. Data, specimens, and future research
  12. Waivers, alterations, and emergency pathways
  13. Ethics review and investigator responsibility
  14. Reading a consent form as a participant
  15. The standard is a real choice
  16. References

Informed consent is a conversation and decision process, not a signature ceremony. If someone invites you into a study, you should come away knowing that the activity is research, what participation involves, which risks and benefits are reasonably foreseeable, which alternatives exist, and that the choice is voluntary.

The form matters because it preserves required information and documents agreement. It cannot cure rushed discussion, unreadable language, or coercion. It cannot cure missing risks, unanswered questions, or lack of decision-making capacity. Consent continues after enrollment as information and circumstances change.

Recruitment materials, advertisements, referral conversations, and screening calls shape expectations before a consent document appears. Claims should be accurate, balanced, and consistent with the protocol and ethics approval.

Language such as “new treatment” can imply proven benefit when the intervention is investigational. Payment should be described as compensation for time and burden, not a reward contingent on completing every visit regardless of withdrawal rights.

The setting matters. A person may feel unable to refuse when the investigator is also a treating clinician, when care is scarce, or when authority relationships are strong. Separating roles, inviting a support person, or using another qualified consent professional can help, depending on the study. Consent should occur with enough time to consider the decision unless a valid urgent pathway applies. A long form handed over moments before a procedure does not create meaningful deliberation.

The core information#

Requirements vary, but common elements include that the activity is research, its purpose, expected duration, procedures, experimental aspects, foreseeable risks or discomforts, potential benefits, alternatives, confidentiality, compensation and treatment for research injury where applicable, contacts, and the voluntary nature of participation.

The discussion should also address practical burdens: visits, travel, and fasting. It should address devices, diaries, and remote monitoring. It should address contraception or pregnancy testing where relevant, restrictions, costs, and what happens after the trial.

Randomization and masking need plain explanation: a participant should understand how assignment occurs, the chance of each group, whether placebo is possible, and what care is available outside the study.

Uncertainty should be honest. “May help” is different from “will help.” Early-phase research may offer little prospect of direct benefit. A trial can be scientifically valuable even when participants receive no personal benefit, but that fact must be clear.

Therapeutic misconception#

Therapeutic misconception occurs when a participant believes the primary purpose of research procedures is to choose what is best for that individual, rather than to answer a study question under a protocol.

Clinical care and research can coexist. Investigators still owe duties of care and safety. Yet randomization, fixed visits, experimental procedures, and data collection may be driven by the protocol rather than personalized need.

Consent should explain this boundary without suggesting that participants are merely data sources. Their welfare remains central, and clinically necessary care should not be withheld outside approved trial conditions. Checking understanding is better than repeating a disclosure, so ask the person to explain in their own words why the study is being done, how assignment works, the main risks, and what choices remain.

Capacity is decision-specific#

Decision-making capacity generally concerns the ability to understand relevant information, appreciate how it applies, reason about options, and communicate a choice. It can vary by decision and over time.

A diagnosis, disability, age, language, or educational level does not by itself establish incapacity, and support can improve decision-making: interpreters, simplified material, visual aids, hearing devices, breaks, and involvement of a trusted person.

When a person lacks capacity under applicable rules, a legally authorized representative may be able to consent. Who qualifies and for which research depends on jurisdiction. The participant's preferences and assent or dissent should still be respected to the extent possible. Capacity can fluctuate with illness, medication, fatigue, or delirium. Research involving progressive cognitive conditions may need plans for reassessment and transition to representative permission.

Children and assent#

Research involving children typically requires parent or guardian permission under applicable rules and the child's assent when the child can provide it, and assent is an affirmative agreement, not mere failure to object.

Information should match developmental level. The child should understand what will happen, what may be uncomfortable, and that questions are welcome.

Rules differ on when one or two parents are required, when permission can be waived, and which risk categories are allowed. Ethics review evaluates these conditions. When a participant enrolled as a child reaches legal capacity during an ongoing study, consent for continued participation may need to be obtained directly, subject to local requirements and study circumstances.

Voluntariness and undue influence#

A voluntary decision is free from coercion and improper influence. Coercion involves a threat of harm used to obtain compliance. Undue influence is an excessive or inappropriate offer or pressure that distorts judgment.

Payment is not automatically undue. Amount, timing, burden, local context, and whether payment is prorated matter. Reimbursement for travel may improve fair access. Withholding all payment after withdrawal can pressure continued participation.

Dependent relationships require care. Employees, students, and prisoners may perceive consequences for refusing. So may military personnel and patients reliant on scarce treatment. Safeguards should address the actual power structure. Consent materials should say that refusing will not remove care or benefits you are entitled to and that withdrawing will not bring a penalty, subject to clear explanations of what follow-up may be medically necessary.

Understandable and accessible communication#

Forms should use concise sentences, defined terms, headings, and formatting that supports reading. Length alone is not protection. Dense legal language can obscure the information people need most.

Translation should preserve meaning and be reviewed through an appropriate process. A family member should not be assumed to replace a qualified interpreter, particularly for complex risk or sensitive information.

Accessible formats may include large print, audio, and captioning. They may include screen-reader-compatible documents, sign-language interpretation, or supported communication. The method should fit the participant without removing substantive content. Health literacy should be addressed respectfully. Teach-back asks the person to explain the plan and reveals what the team needs to clarify; it is not an exam the participant passes or fails.

Electronic consent can include text, video, graphics, questions, and electronic signatures. FDA and OHRP guidance recognizes that electronic methods can satisfy requirements when implemented appropriately.

The system should verify identity as required, present the approved version, and record signatures and dates. It should protect privacy, permit questions, and provide a copy. Audit trails and version control help show which information the participant received.

Remote consent should have a fallback for poor connectivity, inaccessible technology, or need for in-person support. A video module cannot replace a responsive conversation. Cybersecurity and data minimization matter because the consent platform itself collects personal information. Vendor roles, retention, and cross-border transfers should be governed.

Participation lasts beyond the enrollment meeting. New safety information, changes in procedures, new alternatives, or changes in study status may affect willingness to continue.

Some updates require revised written consent approved by the ethics committee. Others may be communicated through a letter or discussion. The pathway depends on importance, urgency, and applicable requirements.

Immediate action may be needed to remove an urgent hazard before prior approval, followed by required reporting; protecting the participant comes first, but emergency action should not become a route around oversight. Routine visits can reinforce consent by asking whether questions arose and confirming continued willingness. This is especially important in long studies or those with evolving burden.

Withdrawal and already collected data#

Participants can stop future research procedures. They may agree to some follow-up while declining the intervention, or decline further contact except safety information. The available choices should be explained.

Data already collected may remain in the study record when needed for scientific integrity, safety, or legal obligations. Rules for identifiable specimens, future research, and data deletion vary.

The consent should distinguish stopping treatment, stopping visits, stopping data collection, and withdrawing permission for optional future use. A single word, “withdraw,” can hide several decisions, which is why investigators should document which one you made and provide clinically appropriate transition or follow-up where it is needed.

Data, specimens, and future research#

Consent for the main trial does not automatically explain every future use of data and specimens. Plans should address storage, coding, and genetic analysis. They should address sharing, commercial use, and return of results. They should address recontact and withdrawal limits where applicable.

Broad consent may permit categories of future research under governance, but it is not unlimited permission, and the scope, review, privacy protections, and whether participants will receive results should be clear.

Genomic data can be difficult to de-identify fully and may have implications for relatives. Returning an individual finding to you requires validity, actionability, confirmation, and a communication plan, so an incidental-findings policy should match what the study can responsibly detect and manage rather than promise comprehensive health screening.

Waivers, alterations, and emergency pathways#

Some minimal-risk research can qualify for waiver or alteration of consent when strict criteria are met and an ethics committee documents the determination. FDA's December 2023 final rule created a limited waiver or alteration pathway for certain FDA-regulated minimal-risk clinical investigations.

Emergency research can have separate exceptions when prospective consent is not feasible, the condition is life-threatening, available treatment is unproven or unsatisfactory, and detailed regulatory and community safeguards are met. Exact criteria differ.

These pathways are not general permission to enroll people who are hard to reach: scientific necessity, risk limits, rights and welfare, practicability, and additional information after participation may all matter. The protocol should identify the legal basis and oversight decision. A public summary should not imply that “consent was unnecessary” without context.

Ethics review and investigator responsibility#

An institutional review board or independent ethics committee reviews the protocol, consent materials, and recruitment. It reviews payment and updates. Approval does not transfer the investigator's duty to conduct a valid process.

The person obtaining consent should be qualified, know the study, answer questions accurately, and recognize when clinical or language support is needed. Delegation should be documented, with investigator oversight.

Monitoring can review whether the correct version was used, signatures and dates are present, consent preceded procedures, and reconsent occurred when required. These checks detect documentation failures. Interviews, observation, or quality review may be needed to assess the process itself.

Deviations should be evaluated for effect on participant rights and safety, reported as required, and corrected. Backdating or recreating consent records is unacceptable.

Mark the procedures that are done only for research. Identify the main risks, the alternatives, and the assignment method. Identify the costs, the payment, the data use, and what happens if you withdraw. Ask who provides care for a research injury, and who can answer your rights questions outside the study team.

Ask for time, a copy, an interpreter, or a support person if any of those would help you. It is reasonable to ask what is still unknown, and what evidence supports the benefit you are being offered.

The GCP revision guide places consent within the full trial-quality framework. Shared decision-making explains a related clinical process with a different purpose. The site's research overview connects ethics and evidence.

The standard is a real choice#

A technically complete form is not the goal. The goal is a decision you can understand, consider, and make voluntarily, with protection continuing after enrollment.

Good consent leaves a traceable record, but its clearest sign is practical: you know what the study is, what will happen, what remains uncertain, what alternatives exist, and that saying no or stopping later remains possible.

References#

  1. FDA informed consent final guidance, August 2023
  2. ICH E6(R3) Principles and Annex 1
  3. World Medical Association 2024 Declaration of Helsinki
  4. HHS Common Rule, 45 CFR 46
  5. FDA explanation of the minimal-risk informed-consent waiver final rule
  6. FDA and OHRP guidance on electronic informed consent

This article is general information and education, not medical or legal advice. Research requirements and participant rights vary by study and jurisdiction.

Questions and answers

Is a signed consent form enough to prove valid informed consent?

No. A signature documents agreement, but valid consent also requires capacity or proper representation, relevant information, understanding, voluntariness, opportunity for questions, and a process consistent with applicable requirements.

Can a trial participant withdraw after signing?

Yes. Participation is voluntary, and a person can stop future trial procedures. What happens to data or samples already collected depends on the consent, protocol, applicable law, and scientific-integrity obligations.

Is electronic consent valid?

It can be when applicable requirements are met, identity and documentation are reliable, information is understandable and accessible, questions can be answered, and the participant receives a copy.

When is reconsent needed?

Reconsent may be needed when new information, protocol changes, new risks, changed participation, or a participant reaching legal capacity could affect willingness to continue, subject to ethics and local requirements.

Can informed consent ever be waived?

Limited waivers, alterations, or emergency exceptions exist in some jurisdictions under strict criteria and ethics oversight. Operational convenience is not enough, and rules differ by study and authority.