For several years, a growing number of clinical trials have tested elements that didn't exist in the classic model: telemedicine visits, medications delivered to patients at home, data drawn from electronic health records instead of measurements collected specifically for the trial. The practice evolved faster than the specific regulation — but these approaches remained, until now, more the exception than the norm, tested case by case, without a detailed framework fully covering them.

On June 3, 2026, this changed officially. The International Council for Harmonisation (ICH) — the international body that sets Good Clinical Practice (GCP) standards, with participation from the US, Europe, Japan, and other regions — adopted Annex 2 to the E6(R3) guideline, the document that, for the first time, provides a complete and detailed framework for decentralized, pragmatic, and real-world-data-based clinical trials.

1. What Annex 2 actually is

Annex 2 isn't a standalone document — it complements the general Principles and Annex 1 (already in effect in the EU since July 23, 2025), explicitly extending them to three modern methodologies:

Decentralized elements: trial activities conducted outside the traditional research center — at the participant's home, through local healthcare centers, or through remote interactions (video calls, mobile apps, wearable sensors).

Pragmatic elements: trial procedures designed to align with usual clinical practice for that condition, rather than being built separately from it — including eligibility criteria and a data collection schedule that reflect how patients are actually cared for, rather than an artificial pattern designed in isolation purely for research purposes.

Real-world data (RWD): information about patients' health status, collected from sources outside the trial itself — electronic health records, registries, insurance claims databases.

The document explicitly states that it does not endorse or recommend any specific methodology — it only provides the framework through which any of them can be applied correctly, if a sponsor chooses to use them, while keeping participant protections and the reliability of results intact. A classic trial, with none of these elements, remains a fully valid option — Annex 2 does not require anyone to decentralize anything.

2. What actually changes for those who choose to use these methodologies

The document sets out detailed rules across three levels — the ethics committee, the investigator at the site, and the trial sponsor. A few practical changes with direct impact:

Informed consent can be obtained remotely — but the investigator must verify the participant's identity (for example, by checking an official ID during a video call), using a verification method specified in advance. The participant must always have the option to choose the classic, in-person or paper-based approach, if they prefer it.

The investigational drug can be shipped directly to the patient's home — by site staff, by the sponsor directly (where local regulation allows), by a home nurse, or by a local pharmacist. Certain requirements remain mandatory: protecting confidentiality, confirming the medication reached the correct person, a clear process for return or destruction, and preserving blinding, where applicable.

Investigator oversight no longer has to be uniform — it can range from direct supervision to simply reviewing essential documents remotely, depending on the actual risk of each activity. In practice, a research site's resources can be allocated differently, not identically, across all procedures. When local healthcare professionals carry out a procedure that is already part of usual clinical practice, clear arrangements for sharing data with the lead investigator are sufficient — full training on the trial protocol is only required if the activity genuinely calls for it.

Responsibility for real-world data remains entirely with the sponsor — even when that data is held by a hospital, a registry, or an insurer, the sponsor must be able to access individual-level data for quality control, and the level of access required increases with how critical that data is to the trial's final outcome.

Using this data also requires a separate consent or permission from the person whose data is being used — distinct from the consent originally given to a doctor for their routine care. For example, if someone visited a doctor years ago and that data is now being used in a new clinical trial, the person must have specifically agreed to this new use. Importantly, this agreement only means their old data can be analyzed by researchers — it does not mean the person has agreed to actively participate in the trial or receive an experimental treatment. These are two entirely different decisions, with different risks, requiring separate consents.

The document specifies that this consent can be obtained in two ways, depending on each country's law and how identifiable the person whose data is being used remains. Either through a broad consent, given in advance, by which the person agrees that their data may be used in future, not-yet-specified research. Or, in certain regions, through approval from an independent ethics committee, which can substitute for individual consent for this secondary use of data.

3. Why it matters, even though no one is required to decentralize a trial

Annex 2 officially becomes part of the GCP standard applicable in the European Union on January 15, 2027. GCP doesn't function as a standalone law — it works by reference: the EU Clinical Trials Regulation (536/2014), which is directly binding law in all member states, requires that any clinical trial comply with good clinical practice, while the CHMP/EMA decides, through official adoption of each version, what exactly that standard means at any given time. In practice, from January 2027, if a sponsor chooses to use decentralized elements, pragmatic elements, or real-world data, exact compliance with Annex 2's rules becomes legally mandatory — not optional or merely recommended.

Regulators in the EU, UK, and Switzerland have already been applying the Principles and Annex 1 since summer 2025. The United States has not yet set an official timeline for adopting Annex 2 — although the standard was developed jointly through ICH, with US participation, Europe is the first major region to give it a firm legal effective date, becoming the first to turn it into a binding requirement.

Annex 2 doesn't force any sponsor to abandon the classic clinical trial model — but for those who choose to use home visits, telemedicine, or existing medical records, the rules to follow are no longer vague or open to interpretation, but clearly defined and, in Europe, legally binding starting January 15, 2027. For any organization involved in clinical trials, the practical question is no longer whether these methodologies are allowed — they have been for a while — but whether internal processes are already aligned with the precise rules they must now follow, if they choose to use them.

Sources:

• International Council for Harmonisation (ICH). E6(R3) Guideline for Good Clinical Practice — Annex 2. Adopted 3 June 2026. https://database.ich.org/sites/default/files/ICH_E6(R3)_Annex%202_Guideline_Step%204_2026_0603_0.pdf

• European Medicines Agency (EMA). ICH E6 Good clinical practice - Scientific guideline. https://www.ema.europa.eu/en/ich-e6-good-clinical-practice-scientific-guideline