ePROElectronic Patient-Reported Outcome

Clinical Outcome Assessments (COAs) are measures of how a patient feels, functions or survives, captured in a trial. COAs come in four flavours by who is doing the reporting: Patient-Reported Outcomes (PRO — the patient), Clinician-Reported Outcomes (ClinRO — the clinician), Observer-Reported Outcomes (ObsRO — a caregiver, parent), and Performance Outcomes (PerfO — a standardised task performed by the patient). When any of these are captured on an electronic device, the prefix 'e' is added — ePRO, eClinRO, eObsRO, ePerfO — and the umbrella term is eCOA.

ePRO is the largest and most discussed segment because it is the data source most often used to support labelling claims that are inherently patient-perspective: pain, fatigue, function, satisfaction, quality of life, symptom burden. FDA's December 2009 PRO guidance remains the foundational document; the more recent PFDD Guidance series (Guidances 1–4, with Guidance 3 on instrument selection and development finalised in June 2022) updates the conceptual framework for modern trials.

FDA's 2009 PRO guidance lays out the qualification expectations for any instrument used to support a labelling claim: content validity (does the instrument measure the concept of interest in the target population?), construct validity (does it correlate with related measures?), reliability (does it give the same result on repeat administration when nothing has changed?), ability to detect change (is the minimum important difference defined?), and equivalence between paper and electronic modes if both are used.

Equivalence is the central historical question for ePRO. If a sponsor migrates a validated paper PRO to an electronic device, FDA expects evidence that the modes produce equivalent measurements — usually through a small equivalence study (cognitive debriefing plus a quantitative equivalence trial) following ISPOR Task Force recommendations. Minor formatting changes (typography, screen layout) typically need only usability testing; meaningful changes (item wording, response options, recall window) require full re-qualification.

Sponsor-provisioned devices were the historical norm — handheld study devices distributed to every participant — because they eliminated cross-device variability. BYOD (the participant uses their own smartphone) has accelerated since 2020 because it removes a logistics burden, removes a digital exclusion vector for participants comfortable with their own device, and is now broadly accepted by FDA when supported by appropriate equivalence and usability evidence.

BYOD-specific risks that the sponsor and ePRO vendor must address: screen-size variability (a question may layout differently on a 5.5" phone than a 6.7" tablet), OS version diversity, network connectivity gaps (data captured offline and synced later — what is the timestamp?), unique identification of the participant (the device is owned, but is the participant the one entering?), and security of the app on a non-sponsor-controlled device.

An ePRO record is an electronic source record under 21 CFR Part 11 when no paper original exists — which is the normal case for a properly designed ePRO. The record must be attributable to the participant, contemporaneous with the patient experience, original (the electronic record is the source, not a transcription), and accurate (the instrument has been validated, the device renders the question as designed).

ICH E6(R3), finalised at Step 4 in January 2025, formalises the GCP expectations for computerised systems used in trials. Section 4 requires sponsors to maintain a documented assessment of every computerised system, scaled to its risk to participant rights, safety and data reliability. ePRO is squarely in scope. The same assessment expectations apply whether the ePRO is delivered on a sponsor device, BYOD, telehealth platform or hybrid.

• A validated instrument library — the vendor has done the equivalence work for the standard instruments your protocol uses (EQ-5D, PROMIS, BPI, etc.) and can show the evidence package.
• Documented Part 11 controls — audit trail, e-signature, RBAC — not 'compliant by design'.
• Offline capture with correct timestamping — participant entry time stored separately from sync time.
• Multilingual support with linguistic validation evidence — translation and back-translation per ISPOR guidance.
• Sponsor-controlled data export with full audit-trail data, not just the summary value per visit.
• A vendor audit report (or SOC 2 + 21 CFR Part 11 audit) that the sponsor's QA can review before contracting.

• Treating ePRO as a feature of the EDC instead of a Part 11–controlled system in its own right.
• Using the paper PRO's printed score-band cutoffs on the ePRO without re-validating equivalence.
• Allowing back-dating — letting participants enter yesterday's pain score 'just to keep the diary complete'. This is a contemporaneity failure and an inspection finding.
• Failing to document the device, OS version and app version associated with each entry — needed to investigate any subsequent rendering question.
• Using sync-time as the primary timestamp on offline-captured data.