The paperless clinical research site: what to digitise first

The phrase eSource covers many setups, from a full electronic medical record feed to a form completed on a tablet during the visit. Capture sits at the lighter end of that range: the investigator or coordinator enters the observation straight into the eCRF, and the participant enters their own reports straight into an ePRO task. Because nothing is transcribed afterwards, one common error class disappears.

Whether that entry is the original record is a protocol decision. Many studies treat the participant’s own diary as the source, and some treat site entries into an eCRF as source for visit-measured values. Others keep the medical record as source and use the eCRF as a transcription target. The FDA’s final guidance on electronic systems, records and signatures in clinical investigations (a questions-and-answers document published in October 2024) is the usual reference for how regulators think about trustworthy electronic records; it is non-binding, so read it with your sponsor and your regulatory contact. Whichever model you pick, write it into the monitoring plan so that source data verification is consistent. In Capture, source data verification can be set per field, which lets a monitor focus on the values that carry the endpoint.

The record-keeping qualities people usually look for are attributable, legible, contemporaneous, original and accurate, sometimes extended by the ALCOA-plus checklist covered on our ALCOA-plus checklist page. A field-level audit trail with timestamp, user identity, old value, new value and reason speaks to attribution and change history, and role-based access keeps one person from using another’s account for another’s work. For the controls in more detail, see the 21 CFR Part 11-aligned EDC page. Compliance remains a shared responsibility between the software and your own validated use of it.

Consent is a process, not a signature. The conversation with the participant is the point; the form records it. A good digital consent keeps that order. The participant reads the document on a tablet or their own phone, the researcher answers questions in person or remotely, and the signature comes last. Our eConsent product page covers the mechanics: the participant signs on screen, confirms with an email one-time code, and acknowledges the legal-binding statement before signing. The investigator or authorised site staff then countersign after re-entering their password, with a meaning statement, and the document counts as fully executed only after that step.

Each signature is cryptographically linked to the exact document content at the moment of signing, so a later edit to the consent text cannot silently apply to an earlier signer. Optional witness co-signature and legally authorised representative signature are supported. The eConsent without paper forms page describes the participant experience, and the consent templates give editable starting texts. Your ethics committee approves the consent wording and the process; confirm how it wants electronic consent described in the submission.