Operationalizing Decentralized Trial Elements: A Site Network Playbook for ICH E6(R3) Annex 2
Published November 2026 — A practical playbook for site networks supporting trials with decentralized elements, organized around the responsibilities that stay with the investigator when trial activities leave the site.
The Activity Moves. The Responsibility Does Not.
Trials with decentralized elements are now routine. Remote visits, home nursing, direct-to-participant shipment, and digital data capture appear in protocols across therapeutic areas, usually in hybrid designs that keep some visits at the site. Until this year, the Good Clinical Practice guideline itself said little about them, and sites relied on regional guidance and sponsor instructions.
The final ICH E6(R3) Annex 2, adopted on June 3, 2026, changes that. It extends GCP to trials with decentralized elements, pragmatic elements, and real-world data sources. Its central message for sites is simple: the investigator remains responsible for the conduct of the trial and the welfare of participants wherever the activity takes place and whoever performs it.
This paper sets out how a site network can turn that principle into working procedures. It describes six operational workstreams, a five-phase implementation sequence, and the questions a site should settle with each sponsor before startup. It is written for site directors, quality leads, and the sponsor teams who work with them.
Six Operational Workstreams
Each workstream corresponds to a responsibility that the site keeps when part of the trial is conducted remotely.
Map every decentralized activity in the protocol to a named role. For each, record who performs it, who they report to, how they were trained, and how the investigator reviews their work. Home health nurses and local clinicians belong on the delegation log when they perform protocol procedures.
Write a remote visit procedure that covers identity verification, privacy at the participant's location, what is assessed and how it is documented, and the trigger for converting to an in-person visit. Source documentation for a telehealth visit should be as complete as for a clinic visit.
Agree with the sponsor on each step of direct-to-participant supply: who releases the shipment, how temperature is monitored, who confirms receipt and condition, how dosing is recorded, and how unused product is returned and reconciled. The site's accountability records should show the full chain.
List every system that generates trial data, including eConsent, ePRO, wearables, home health records, and the EHR. For each, identify what counts as source, who can see it, how the site receives it, and how long it is retained. Unreviewed data streams are the most common gap.
Define who looks at each incoming stream for safety-relevant information, within what time, and how a concern reaches the investigator. A reading from a device or a note from a home visit should not wait for the next scheduled visit to be seen.
Annex 2 emphasizes participant-centered design. Give participants one point of contact at the site, plain instructions for technology and shipments, and a clear way to ask for an in-person visit. Decentralized elements should lower the burden of taking part, not move it to the participant.
Implementation Sequence
A network does not need to build everything at once. The sequence below starts from current practice and moves toward a common standard that holds across sponsors and sites.
Assess: Inventory Current Practice
Start with the trials the network is already running. List each decentralized activity, the procedure that governs it, and where that procedure came from. Most networks find a mix of sponsor-specific instructions and local habits. The inventory shows where practice differs from site to site for the same activity.
Standardize: Write Network Procedures
Draft one network procedure for each workstream above, written so it can be applied across sponsors. Sponsor-specific requirements then become additions to a common base. A coordinator who moves between studies should find remote visits and home shipments handled the same way each time.
Train and Qualify: People and Vendors
Train site staff on the procedures and document it. Apply the same expectation to third parties who perform trial activities for the site. Where the sponsor contracts the vendor, confirm in writing how the investigator will oversee the vendor's personnel and receive their records.
Verify: Test Before First Participant
Before a study opens, walk through one complete remote visit and one mock shipment from release to reconciliation. Check that each record lands where the procedure says it should. Problems found in a dry run cost far less than a deviation found at monitoring.
Monitor: Review and Improve
Track a small set of indicators for decentralized activities: remote visits completed as planned, shipments with temperature excursions, time from data receipt to review, and deviations linked to remote procedures. Review them at a set interval and adjust the procedures when a pattern appears.
Questions to Settle with the Sponsor Before Startup
Most problems with decentralized elements trace back to something left undefined at startup. These six questions cover the usual gaps.
Ask the sponsor to mark each assessment in the schedule of activities as on-site only, remote permitted, or remote preferred. Ambiguity here is the root of most remote visit deviations.
Whether the sponsor or the site contracts home health and courier services changes who qualifies the vendor and who receives its records. The investigator's oversight obligation is the same either way.
Agree before startup where the original record lives for each remotely collected data point, and how the site will access it during monitoring and inspection.
A dropped video call, a device that stops transmitting, or a participant without connectivity should each have a documented fallback that preserves the visit window where possible.
Check that the consent describes home visits, shipments, devices, and data sharing as they will actually happen at this site. Generic language written for all sites often misses local practice.
Annex 2 takes effect in the EU on January 15, 2027, and the FDA has not yet published its final version. A multi-region protocol should be designed to the final ICH text from the start.
Why a Network Has an Advantage
An independent site meets decentralized elements one sponsor at a time and has to absorb each sponsor’s way of doing things. A governed network can set one standard and apply it everywhere. Procedures are written once, training is shared, vendor qualification is reused, and what one site learns from a difficult study reaches the others.
For sponsors, that consistency reduces the variability that makes hybrid trials hard to monitor. When every site documents a remote visit the same way, central monitoring can compare sites meaningfully and direct attention to real outliers.
Conclusions
Annex 2 does not ask sites to do something new so much as to show that what they already do remotely meets the same standard as what they do in the clinic. The work is in making responsibilities explicit: who performs each activity, who oversees it, where the record lives, and who reviews it.
Networks that settle those questions in common procedures before the EU effective date of January 15, 2027 will be in a position to take on hybrid and decentralized protocols without rebuilding their approach for each one. This paper is a summary for planning purposes and should be read alongside the guideline text and applicable regional guidance.
Planning a Hybrid or Decentralized Protocol?
Talk with our team about how network sites support remote visits, delegated providers, and direct-to-participant supply.