ICH E6(R3): What Changed From R2, and What It Means for You

ICH E6 is Good Clinical Practice: the standard for designing, conducting, recording and reporting trials involving human subjects. R3 is the current revision, and it restructures the guideline rather than amending it.

How much a sponsor has to change depends on how it applied R2: less than the size of the document suggests if R2 was applied with judgment, considerably more if it was applied as a checklist.

Changes from R2

A principles-first structure. R3 leads with overarching principles and then sets out responsibilities. The principles are meant to be applied to circumstances the guideline does not enumerate, because R2 was increasingly being read as an exhaustive list.

Quality by design. Quality is to be built into trial design rather than inspected in afterwards. That means identifying the factors critical to reliability of results and safety of participants at the design stage, and concentrating effort there.

Critical to quality factors. You are expected to identify the factors that matter in your trial and focus monitoring and controls on them, rather than applying uniform effort to everything.

Proportionality. Controls should be proportionate to the risks. A low-risk, low-complexity trial is not expected to carry the apparatus of a large multi-regional one.

Data governance. R3 assumes trials run on electronic systems and decentralized elements, and sets data integrity expectations across the data lifecycle rather than assuming paper.

An annex structure. R3 is built to be extended by annexes, which is how ICH intends to address specific trial types without reopening the core guideline.

Gaps against R3

If your quality system already reflected risk-based monitoring and data integrity expectations, the gap is mostly documentation: showing that you identified critical to quality factors and that your controls follow from them.

Where it bites:

• SOPs written as procedure without rationale. R3 expects proportionality, which requires you to be able to say why a control exists.
• Uniform monitoring. 100% source data verification everywhere is the opposite of what R3 asks for, and it was already the opposite of what R2 asked for after the addendum.
• Vendor oversight. Sponsor responsibility for delegated activities is unchanged and stated more directly.
• Computerized systems. Expectations around validation and access control apply to the systems holding trial data.

Doing an R2 to R3 gap assessment

A clause-by-clause mapping is less useful than answering these questions:

1. Can you point to your critical to quality factors for each ongoing trial?
2. Do your monitoring and control decisions follow from those factors, in writing?
3. Are your computerized systems governed, with access control and audit trails that hold?
4. Is your vendor oversight documented as oversight, not as contracting?

A sponsor that can answer yes to each can usually show compliance. One that cannot has a gap, however thick the SOP manual is.