ICH GCP E6 completed
Where does Annex 2 sit in the E6(R3) structure?
E6(R3) is built in three layers. In the first layer, the principles set the overarching ethical and scientific framework.
Annex 1 as the second layer operationalizes those principles for conventional interventional trials, covering the work of ethics committees, investigator and sponsor responsibilities, data governance, the protocol and more.
The third layer, Annex 2, extends that operational guidance specifically to non-traditional trials. Annex 2 must always be read together with Annex 1 and the principles, not instead of them.
Two principles run through every section of ICH GCP E6(R3) Annex 2.
Fitness for purpose: every methodology, data source, and technology used in a trial must be demonstrably appropriate for its intended use, it should not be used simply because it is fancy, available or convenient.
Proportionality: the level of oversight, quality control, and data access should always match the actual risk to participant safety and data reliability, not be applied uniformly regardless of context.
These two principles are not background philosophy. They are the practical test that every design decision in a non-traditional trial must pass.
ICH GCP: The problem Annex 2 solves
Decentralized and pragmatic trial elements in clinical trials are not new. Sponsors have been shipping investigational products to patients’ homes, collecting data via wearables, running remote visits, and pulling outcomes from electronic health records for years – accelerated sharply by COVID-19. But under ICH E6(R2), none of this had explicit ICH GCP backing. Teams operated in a sometimes grey zone, piecing together concepts from FDA guidance documents, EMA reflection papers, CIOMS reports, guidelines of national competent authorities, handouts of ethics committees. Auditors had no shared global standard. Sponsors had no common language to document their decisions.
Annex 2 reduces the size of the grey zone. It does not invent new principles; it makes the existing ones travel into the settings where many clinical trials now actually happen.
The three “methodologies” defined
🏠 Decentralized elements: trial-related activities conducted outside the investigator’s location – home visits, local healthcare centers, mobile units, video calls, and digital health technologies (wearables, apps, sensors) used to perform procedures and collect data remotely.
🩺 Pragmatic elements: integration of usual clinical practice into trial design – real-world eligibility criteria, streamlined data collection aligned to routine care, healthcare professionals performing trial activities as part of their normal working day.
🌐 Real-world data: patient health data collected outside clinical trials – electronic health records, claims databases, registries, repurposed as trial endpoints, outcome data, or external controls.
What ICH GCP E6(R3) Annex 2 actually requires
1. Informed consent – some more content for the data protection section
Consent materials and processes must be tailored to the specific design elements used in the trial. In the practice this means to blow up the data protection part of the ICF template.
While Annex 2 encourages the use of remote consent, sponsor have to ensure compliance with regulatory requirements and the mind set of ethics committees. When used, the investigator must verify the participant’s identity before consent is obtained, for example by checking an official ID document via video call. Crucially, both the verification method and the data privacy safeguards must be pre-specified in the protocol.
A paper-based or in-person alternative must remain available to the extent feasible. Annex 2 explicitly recognizes that not all participants are equally comfortable with digital systems.
The consent materials themselves must go further than in conventional trials: they must describe exactly what participant data will be collected, how it will be used, and which parties will have access to the participant’s personal information, including their home address and health records. This is particularly important when trial activities are conducted at the participant’s home or data are collected via digital health technologies, since participants may not realize the extent of data sharing involved.
When a clinical trial uses real world data, it has to be ensured that their use is enabled by an approved consent, which in some cases is obtained outside the context of the clinical trial.
2. Direct shipment of investigational products - with conditions and with documented responsibility
Both the investigator/institution and the sponsor may arrange direct shipment of the investigational product to the participant’s home, if this permitted by local regulatory requirements, authorities, and ethics committees. If permitted, this is a significant practical enabler for decentralized trials. But it comes with specific, non-negotiable conditions.
When the investigator arranges shipment, the following must be addressed:
(a) the participant’s privacy and disease status must be kept confidential;
(b) the product must be received by the intended recipient — the participant or an appropriate designee (not any neighbor and not any member of the household);
(c) full accountability must be maintained for receipt, storage, handling, administration, return, and destruction;
(d) blinding must be protected where applicable; and
(e) participant support tools must be provided such as tutorials, brochures, visual aids, and contact details for technical support.
When the sponsor arranges shipment, the same conditions apply. Additionally, shipment must follow the protocol and may only be initiated after explicit authorization by the investigator.
The investigator/institution retains clinical responsibility for the safe and appropriate use of the product regardless of who organized the logistics. Roles and responsibilities must be clearly documented, and predefined communication procedures must be in place between the sponsor, investigator, and any service provider involved in shipping.
Before any shipment arrangements are made, the sponsor must assess the approach during protocol development, considering product stability, storage requirements, reconstitution complexity, route of administration, trial population, safety profile, need for in-person post-administration observation, and the need for emergency plans such as rescue medication availability.
3. Investigator oversight - proportionate, and pre-designed
Investigator oversight in non-traditional trials might not require physical presence. Annex 2 is explicit: the appropriate level ranges from direct supervision to remote communication (video, telephone, email) to review of essential records only. What determines the level is not convenience but the nature of the activity, the criticality of the data being collected, and the risk to participant safety.
A key scenario addressed specifically in Annex 2 is where healthcare professionals perform trial-related activities as part of usual clinical practice, for example, a GP conducting a routine visit that also serves as a trial assessment. Annex 2 clearly states:
- If the activity does not require protocol knowledge, the investigator does not need to delegate formally but must have arrangements in place ensuring data are shared, records are retained (e.g. as certified copies), data privacy is maintained, and data integrity is ensured.
- If the activity does require protocol or IB knowledge, it must be performed by an appropriately trained, formally delegated healthcare professional.
In both cases, the level of oversight should not be low if the data collected are critical for the safety of participants or for the reliability of the trial. Regardless of whether the healthcare professional was arranged by the sponsor or the investigator, the investigator remains responsible for ensuring the resulting records meet protocol requirements.
4. Real World Data - a fitness test, not just a data agreement
ICH E6 (R3) Annex 2 does not usher in the replacement of interventional clinical trials by studies that collect Real World Data only. Annex 2 does not embrace the multi-faceted opportunities for the use of Real World Data that ICH M14 and (currently emerging) ICH E23 vividly illustrate. Instead, it handles Real World Data with care. While it actually opens interventional clinical trials for them, it appears to be somehow not sure what for. It refers to “various roles” Real World Data might play, “including but not limited to endpoint or outcome data or serving as an external control”.
In line with this uncertainty, Annex 2 asks Sponsors to demonstrate that Real World Data are actually fit for purpose – meaning both reliable (accurate, complete, traceable, with documented provenance) and relevant (the right data elements to answer the specific trial question with the specific method). For key efficacy or safety endpoints, access to source records is required for quality assurance and quality control, a system-level audit trail is not sufficient. For lower-criticality uses such as exploratory endpoints, proportionate process-level assessments may suffice. Where data from multiple sources are linked, the methodology must be pre-specified in the protocol, including how inconsistencies will be resolved.
5. Safety reporting - multi-source, actionable, by design
In conventional trials, safety data flows from a defined set of site visits. In decentralized and pragmatic trials, it arrives from multiple directions simultaneously: home nursing reports, remote video consultations, continuous digital health technology streams (wearable heart rate, activity, adverse event alerts), entries in electronic health records, and in-person visits.
The responsibility is shared. The investigator must review and assess health-status information across all these sources, not only data collected at the site. The sponsor must ensure that appropriate processes and procedures are in place so that this information is accessible to the investigator in a timely manner, in accordance with the protocol.
Annex 2 adds a specific and important requirement for digital health technology data: it must be provided to the investigator in a way that is relevant, meaningful, and manageable. This is deliberate language. A raw data export from a wearable – thousands of data points with no clinical interpretation – does not meet this standard. The sponsor is responsible for designing an information pipeline that enables the investigator to form an effective overview of each participant’s health status and make timely, appropriate medical decisions. That pipeline must be described in the protocol before the trial starts.
The entire approach to safety management, including how potential safety concerns will be identified, communicated to investigators, and acted upon, must be described in the protocol or a dedicated safety management plan, taking into account the specific methodologies used in the trial.
Summary
ICH E6 (R3) Annex 2 does not give permission to run trials with non-traditional elements differently than the well-known clinical trials. It gives a framework for doing what many teams were already doing — properly, consistently, and with a shared ICH GCP foundation. ICH E6 (R3) Annex 2 defines some basic principles for handling decentralised elements, pragmatic elements, Real World Data in the context of quite traditional interventional clinical trial trials with medicinal products. However, if you rely on ICH E6 (R3) Annex 2 only, you might get led astray. E.g., some local authorities might not approve the shipment of medicinal products to the homes of participants. Spanish inspectors will probably use the AEMPS “Guía para la realización de elementos descentralizados en ensayos clínicos”. Like ICH E6 (R3) Annex 1, it is important to read and implement Annex 2 in the context of local legislation, regulatory requirements, ethics committee approvals.
Who needs to act now
If your trials involve any of the following – home delivery of investigational products, eConsent, remote monitoring, wearables, electronic health record-based endpoints, registry data, pragmatic design elements, or healthcare professionals performing trial procedures as part of routine care – ICH E6 (R3) Annex 2 applies to you.
If you have any questions, FGK is your best place to turn to, we will guide you successfully through the jungle!
About the Author:
Prof. Dr. Hermine Wenzlaff
Director General Management Berlin Office
Hermine has been heading FGK’s Berlin office since 2020. She served for many years as a professor of clinical research at the Medical School Berlin, is an expert in ICH-GCP, and is always happy to share her in-depth knowledge with colleagues and interested industry professionals.
FAQs about ICH GCP E6(R3) Annex 2
What is the difference between Annex 1 and Annex 2 of ICH GCP E6(R3)?
Annex 1 defines the core Good Clinical Practice (GCP) principles that apply to all clinical trials. Annex 2 builds on these principles by providing additional guidance for decentralized, pragmatic, and other innovative trial designs. It addresses the use of digital technologies, remote trial activities, and real-world data without changing the fundamental GCP requirements.
How does ICH GCP E6(R3) Annex 2 affect sponsors and clinical trial sites?
Annex 2 provides clearer guidance for modern trial designs. Sponsors should review their risk-based quality and oversight processes to ensure new technologies and external data sources are properly validated and documented. Study sites should clearly define responsibilities for remote activities and data management while maintaining participant safety and data integrity.
When should organizations start preparing for the implementation of ICH GCP E6(R3) Annex 2?
Organizations should begin reviewing their processes as early as possible. Assessing quality systems, digital technologies, vendor oversight, and risk-based procedures in advance will help ensure a smooth transition and favorable review by regulatory authorities.