Types of eSource: Direct Data Capture, EHR, EHR-to-EDC and ePRO

The four types of eSource in clinical trials are direct data capture, the electronic health record used as source, EHR-to-EDC transfer, and electronic patient-reported outcomes (ePRO). Each type records the original value electronically at first capture. The types differ in who makes that capture, what data it holds and how the value reaches the case report form (CRF).

This guide sets out what each type captures, who enters the data and what the regulators say about it. It sits under our complete guide to eSource in clinical trials for UK research sites, which covers the wider record lifecycle, regulation and buying questions.

The type of eSource is decided by whoever touches the record first.

What Separates One Type of eSource from Another?

Three questions separate the types. The US Food and Drug Administration (FDA) frames the first through the data originator: the person, computer system, device or instrument authorised to enter, change or transmit a data element. Its guidance on electronic source data in clinical investigations lists five ways a data element can reach the electronic CRF (eCRF).

• Who is the data originator: delegated site staff, the clinical care team, an interface or the participant?
• What data does the type hold: protocol assessments, routine care data or the participant’s own report?
• How does the value reach the CRF: by direct entry, by an automated transfer or by retyping?

Source data itself is defined in our guide to what counts as source data in a clinical trial. This guide covers only the electronic forms that source takes.

What Is Direct Data Capture?

Direct data capture is eSource entered by site staff into a trial application at the point of care, so the entry is the original record. The European Medicines Agency (EMA) described it in its 2019 qualification opinion on eSource direct data capture, requested by Novartis, as “an electronic application and/or device that allows direct entry of source data” by investigator site staff.

The opinion accepted DDC for trials supporting marketing authorisation applications, on four conditions:

• The system is configured to meet local legal requirements and ICH GCP.
• Data validation and testing are complete before the system is deployed.
• The system is secure and properly maintained.
• The investigator keeps continuous control of the data during and after the trial.

The same opinion states that “only protocol-mandated source data should be transferred and accessible to the sponsor”. DDC therefore captures the assessments the protocol asks site staff to perform.

The EMA guideline on computerised systems and electronic data in clinical trials adds a planning duty: “The protocol should identify any data to be recorded directly in the data acquisition tools and identify them as source data.” DDC gives the site the most control of the four types. It also gives the site the most to maintain, because every form is built from the protocol and validated before first use. Our guide to computerised system validation under GAMP 5 covers the evidence behind that step.

When Is the Electronic Health Record the Source?

The electronic health record (EHR), called the electronic patient record (EPR) in most NHS trusts, is the source when trial data is first recorded there as part of routine care. The clinical team enters the value for the patient’s care. The research team reads it later for the trial.

Common EPR-sourced data at a UK site includes:

• Medical history and diagnoses recorded in clinic.
• Concomitant medications on the prescribing record.
• Routine laboratory results filed from the hospital pathology system.
• Weight, height and observations taken by ward or clinic staff.

The research team has the least control over this type. The trust owns and configures the EPR, and the FDA states in its 2018 guidance on using EHR data in clinical investigations that it “does not intend to assess compliance of EHR systems with 21 CFR part 11”. The value usually reaches the CRF by retyping. The FDA’s 2013 guidance treats retyping from an electronic record as transcription, the same as retyping from paper, and the EMA guideline expects risk-based controls such as double data entry or data monitoring for transcribed data.

Also Read: What Is EDC in Clinical Trials?

How Does EHR-to-EDC Transfer Differ from Using the EHR as Source?

EHR-to-EDC transfer keeps the EHR as the source and replaces the retyping step with a validated electronic interface. The FDA’s 2013 guidance states that for data transmitted directly from an EHR to the eCRF, “the EHR is the source”. The clinical team remains the data originator, and the interface moves mapped values into the sponsor’s electronic data capture (EDC) system.

The FDA’s 2018 guidance describes three levels of connection between the two systems:

The interface carries its own validation duty. The FDA guidance states that sponsors should ensure the interoperability of EHR and EDC systems “functions in the manner intended in a consistent and repeatable fashion”.

Transfer suits structured, coded fields best. Laboratory values, vital signs, demographics and medications are the usual candidates for mapping. Investigator judgements such as adverse event causality often sit outside the EHR, so they need another type of eSource.

What Does ePRO Capture and Who Enters It?

Electronic patient-reported outcomes (ePRO) are eSource entered by the trial participant on an electronic device. The FDA’s 2013 guidance states that when a participant uses a PRO instrument to transmit data directly to the eCRF, “the subject is the data originator and the eCRF is the source”.

ePRO sits within the wider group of electronic clinical outcome assessments (eCOA). eCOA also covers outcomes rated by a clinician, outcomes reported by an observer such as a parent or carer, and performance tests the participant completes. Typical ePRO data includes:

• Daily symptom or pain diaries completed at home.
• Validated quality of life questionnaires completed at a visit or remotely.
• Medication adherence and event diaries.

Devices come in two forms. The sponsor or site can provision a device, or the participant can use an application on their own phone, which the EMA guideline calls bring your own device (BYOD). The guideline expects data captured directly on devices to carry metadata such as “device version, device identifiers, firmware version, last calibration, data originator, timestamp of events”.

The investigator’s duty still reaches this data. ICH E6(R3) makes the investigator responsible for timely review of relevant data from external sources that can affect eligibility, treatment or safety. It names “electronic patient-reported outcome (ePRO) data” among those sources, where appropriate. Our guide to what ICH E6(R3) and the new UK CTR require of research software covers the wider duties that took effect on 28 April 2026.

How Do the Four Types Compare on Control and Coverage?

The four types trade site control against breadth of data. The figure below positions each type on those two axes as an indicative guide, drawn from how each type is defined.

• Direct data capture gives the site control across protocol assessments, which brings the build and validation effort with it.
• EHR as source offers broad routine coverage, in a record the trust configures and controls.
• EHR-to-EDC transfer removes retyping, and its reach stops at the fields the interface maps.
• ePRO records the participant’s own account, and its scope is what the participant can report.

Audit trail quality follows control. Entries the site makes itself carry its own evidence of who entered each value and when, the attributable and contemporaneous attributes of ALCOA+. Entries made in another organisation’s system depend on that system’s trail and on the site’s access to it.

Why Do Industry Groupings of eSource Differ?

Groupings differ because each author sorts eSource by its own concern. The FDA sorts by capture method, the EMA qualified one type on request, TransCelerate sorts by data stream, and a site sorts by who makes the entry.

The TransCelerate grouping appears in a 2020 paper in Therapeutic Innovation and Regulatory Science. Its non-CRF category covers third-party data such as central laboratory results sent to the sponsor. Its devices and apps category covers wearables that collect continuous readings. A UK site usually meets these streams as data it reviews, since the sponsor’s vendor holds the original.

How Many Types of eSource Does a Single Study Use?

Most studies at an NHS site use two or more types at once. A hypothetical oncology visit at a UK trust shows the mix:

1. The research nurse records vital signs and the adverse event assessment on a DDC form.
2. Haemoglobin arrives in the EPR from hospital pathology, and an interface sends it to the sponsor’s EDC system.
3. Concomitant medications sit on the EPR prescribing record, and a data coordinator retypes them into the EDC system.
4. The participant completes a fatigue questionnaire on a site tablet in the waiting room.

Each value has one type and one declared source. The site records that declaration per data point in a source data location log, filed in the investigator site file. The mix also sets the site’s workload: every retyped value is a transcription to check, and every interface is a system to validate.

The mix of types a site runs decides how much of its own record it controls. AQ is launching eSource soon as part of the AQ platform. Book a live demo to see the AQ platform today.