Foundations
ICH E6(R3) restructured GCP around eleven overarching principles — consolidated down from thirteen under the older E6(R2), with two genuinely new additions. Every trial, everywhere, is meant to trace back to these.
ICH E6(R3)
These are interdependent — E6(R3) is explicit that they should be considered as a whole to assure ethical trial conduct and reliable results, not treated as a checklist to satisfy individually.
Every trial has to trace back to the ethical principles set out in the Declaration of Helsinki, applied alongside GCP and whatever regulatory requirements apply locally. The core commitment underneath this: a participant’s rights, safety, and well-being take priority over the interests of science or society, full stop, even when a trial’s scientific value is genuinely high.
This principle is why a trial can be scientifically well-designed and still be shut down or redesigned if it turns out to expose participants to disproportionate risk — good science is a necessary condition for an ethical trial, never a sufficient justification on its own for compromising participant welfare.
Consent is a process, not a signature captured once at enrolment. It has to be voluntary, genuinely understood (not just formally disclosed), and take into account the specific trial’s design, its realistic benefits and risks, the setting it’s conducted in, and increasingly, the technology used to obtain it — R3 explicitly makes room for eConsent and remote consent processes. Emergency situations have their own defined pathway: consent from the participant or their legally acceptable representative as soon as practically possible, following a process the ethics committee has already approved in advance.
A consent form that’s technically complete but that a participant didn’t actually understand doesn’t satisfy this principle — comprehension is the real bar, not just documentation, which is why consent materials are expected to be genuinely clear and concise rather than legally exhaustive.
A trial needs a favourable opinion from an Independent Review Board or Independent Ethics Committee before it starts — a neutral third party has to independently agree the anticipated benefits justify the risks, rather than the sponsor and investigator alone making that judgment. R3 adds an explicit expectation that the IRB/IEC has enough documentation to assess whether the trial is actually operationally feasible, not just ethically sound on paper.
The operational-feasibility addition is a real, practical shift — an ethics committee approving a trial design that turns out to be unworkable at actual sites doesn’t serve participants any better than approving one with a genuine ethical flaw, so R3 asks reviewers to weigh both.
A trial needs adequate nonclinical and clinical information behind it before it starts, and it has to be built on a scientifically sound protocol capable of actually answering the question it sets out to ask. Poorly designed trials aren’t just wasteful of participants’ time and risk exposure — they’re considered unethical in their own right under this framework, independent of how they’re conducted.
This is the principle that connects most directly to E8(R1)’s quality-by-design thinking — a trial’s scientific soundness is treated as a prerequisite for its ethics, not a separate consideration weighed afterward.
Medical decisions and care within a trial have to be provided by appropriately qualified physicians, and more broadly, everyone performing a trial-related task needs to be qualified for that specific role by education, training, or relevant experience — not just qualified in a general sense.
The "specific role" framing matters: someone qualified to run laboratory assessments isn’t automatically qualified to make clinical judgment calls about a participant’s eligibility, and this principle expects trial teams to be explicit about matching individual qualifications to individual responsibilities, not just to competence in general.
Quality has to be designed into a trial from the start, not inspected in after the fact. This means identifying the factors genuinely critical to participant safety and result reliability during planning, and building the trial’s processes around protecting those specific factors, rather than applying uniform scrutiny to everything regardless of how much it actually matters.
This is E8(R1)’s critical-to-quality-factors concept, restated as a GCP principle in its own right — a sign of how central that quality-by-design thinking became to the whole R3 restructure, not just a footnote borrowed from a related guideline.
One of the genuinely new principles — no direct equivalent existed under E6(R2). Trial processes and risk-mitigation strategies should be proportionate to the importance of the data being collected and the actual risk to participant safety and result reliability, and trial designs should be operationally feasible without unnecessary complexity.
This is the principle that formally licenses risk-based, targeted monitoring over exhaustive, uniform monitoring of every single data point — a low-risk observational element of a trial doesn’t need the same intensity of oversight as a high-risk intervention, and treating them identically isn’t more rigorous, just less efficient.
Trials have to be conducted according to a protocol that’s received a favourable ethics opinion, with any amendments going back through that same review process before being implemented. The protocol is the operational contract the whole trial is built around.
This is also where risk-proportionate thinking shows up operationally — protocol deviations aren’t treated as uniformly serious; a deviation that threatens participant safety or data reliability gets a different response than one that doesn’t, even though both are technically departures from the approved plan.
Data has to be recorded, handled, and stored in a way that allows it to be accurately reported, interpreted, and independently verified later. This is the foundation of source data standards familiar from GCP training generally — attributable, legible, contemporaneous, original, and accurate records.
Reliable results and quality (Principle 6) are closely linked but distinct: quality is about how the trial is run, reliability is specifically about whether the resulting data can be trusted once the trial is over — a well-run trial with poor data-handling standards can still fail this principle.
The second genuinely new principle under R3, with no direct E6(R2) equivalent. The roles and responsibilities of sponsors, investigators, and other trial parties need to be clearly defined and understood, with accountability structures that actually match who is responsible for what in practice.
This principle exists because ambiguity about who owns a given decision or oversight task is a real, recurring source of trial failures — not dramatic misconduct, just gaps where everyone assumed someone else was covering something. Making roles explicit is a direct response to that pattern.
Investigational products have to be manufactured and handled according to Good Manufacturing Practice, used strictly in accordance with the approved protocol, and tracked through proper storage and chain-of-custody procedures throughout the trial.
This principle is where GCP and GMP genuinely intersect — a trial’s clinical conduct can be flawless, but if the investigational product itself wasn’t handled to GMP standards, the reliability of what the trial is even testing comes into question.
Worth knowing
Two of these principles — Risk Proportionality and Roles and Responsibilities — have no direct equivalent under the older E6(R2). They're genuinely new, not renamed versions of existing ideas, which makes them worth naming specifically if GCP comes up in an interview: it signals you know the R3 restructure, not just "GCP" as a general concept. The full guideline text lives on the GVP & ICH Guidelines page, alongside the official source link.