ePRO common mistakes
ePRO · Understand the traps and how to avoid them.
Treating ePRO as a general clinical-trials or GCP exam and spending most study time on informed consent, monitoring, and site-initiation procedures, then answering ePRO items with generic clinical-operations reasoning.
Candidates see 'clinical trial' context in the item stem and default to the broad GCP knowledge they already have, because ePRO questions are almost always embedded in a trial scenario. They underestimate how much of the exam turns on instrument design, measurement properties, and data-flow specifics rather than trial administration.
What to do instead
Build a one-page map of what ePRO actually covers - instrument selection and validation, mode effects and equivalence, recall period and response scales, device and platform requirements, compliance and reminder design, data provenance and audit trails, and regulatory guidance specific to COA endpoints - and deliberately separate it from generic GCP content. For every practice item, ask 'which ePRO-specific concept is being tested?' before answering.
Choosing an instrument because it is widely used or familiar rather than because it matches the trial's endpoint, population, recall period, and measurement model - for example picking a generic health-status questionnaire for a symptom-specific endpoint.
Candidates recognize famous instruments by name and assume recognition equals appropriateness. They also conflate 'validated' with 'validated for this use', forgetting that validation is context-specific (population, language, mode, endpoint).
What to do instead
For every instrument question, run a four-part check: what construct and endpoint is the trial measuring, in which population, over what recall window, and in which administration mode. Reject any option that fails one of these, even if the instrument is well known. Memorize the distinction between generic, condition-specific, and domain-specific instruments and when each is defensible.
Assuming that moving a validated paper questionnaire to a screen is automatically equivalent, and ignoring the need for mode-equivalence evidence, cognitive debriefing of the electronic version, or documentation of migration.
The surface content looks identical, so candidates conclude the measurement is identical. They overlook that layout, scrolling, item-by-item presentation, skip logic, and touch input can change interpretation and responses, and that regulators expect evidence of equivalence rather than assertion.
What to do instead
Learn the standard migration pathway: permission from the instrument developer, a rigorous migration process, cognitive interviews with patients on the electronic form, and a usability or equivalence study as appropriate. When an item asks whether electronic administration is acceptable, look for the option that requires documented equivalence evidence, not the one that says the content is unchanged.
Confusing compliance with data quality - treating a high completion rate as proof of good data, or treating a low completion rate as automatically invalidating the dataset, without considering who is missing, when, and why.
Compliance is the easiest ePRO metric to quantify and report, so candidates anchor on the percentage. They forget that non-random missingness (for example, patients skipping entries when symptoms worsen) can bias results even at high overall compliance, while a lower rate with well-characterized missingness may be interpretable.
What to do instead
Always ask about the pattern and mechanism of missing data, not just the rate. Distinguish item-level from visit-level missingness, and distinguish missing completely at random from missingness related to patient status. Practice items that give a compliance figure plus a description of who dropped out, and choose the answer that addresses bias and pre-specified analysis rather than the one that simply praises or condemns the number.
Misreading the recall period and response scale in an item - for example applying a 'past 7 days' instrument to a daily diary endpoint, or interpreting a numeric rating scale as if it were a categorical severity label.
Candidates skim the instrument description to get to the question and carry over assumptions from instruments they know. Recall period and scale anchors are usually stated once in the stem and are easy to lose, and numeric versus verbal rating scales invite intuitive but wrong interpretation.
What to do instead
Before answering, extract and write down three things from the stem: the recall window, the response format (numeric rating, verbal rating, visual analogue), and the direction of scoring (higher = better or worse). Check that the proposed analysis, endpoint definition, or schedule of assessments is consistent with all three. Drill items where the recall period is the deciding detail.
Ignoring the device and platform layer - assuming a web or app-based PRO is interchangeable with a provisioned handheld device, and overlooking connectivity, offline capture, timestamp source, and device provisioning requirements.
Candidates focus on the questionnaire content and treat the technology as a neutral container. In practice, bring-your-own-device versus provisioned device, offline entry with later sync, and server versus device timestamps materially affect data integrity and auditability, and exam items test exactly these distinctions.
What to do instead
Study ePRO systems as a data-capture chain: patient entry, local storage, transmission, server receipt, and audit trail. Learn what must be captured at each step (source of timestamp, confirmation of transmission, handling of offline entries) and why server-side timestamps matter for proving when data were recorded. For any technology item, identify which link in the chain is being questioned.
Treating the ePRO data as automatically equivalent to source data and failing to distinguish the patient's original entry from later edits, derived scores, and transferred datasets when asked about source documentation and audit trails.
In paper trials the source is a physical form with visible corrections; in ePRO the audit trail is invisible unless you know to look for it. Candidates assume the database value is the source and miss questions about original entries, edit history, and who changed what.
What to do instead
Learn the principle that the patient's original electronic entry is the source record, that the system must retain an immutable audit trail of changes, and that derived or transferred data are not source. Practice distinguishing 'original entry', 'audit trail', 'derived score', and 'analysis dataset' in item wording, and choose answers that preserve the original record rather than overwriting it.
Answering regulatory and guidance questions from general impression rather than knowing which document governs which issue - for example mixing up endpoint-qualification expectations, PRO guidance on instrument selection, and computerized-system requirements for records.
Several overlapping guidance documents touch ePRO, and candidates remember that 'guidance exists' without attaching specific requirements to specific documents. Under time pressure they pick the plausible-sounding option instead of the one consistent with the governing framework.
What to do instead
Build a short crosswalk: for each major issue (instrument selection and validation, electronic migration, endpoint claims, computerized records and signatures, data integrity), note which type of guidance or regulation addresses it and what the core requirement is. Answer guidance items by matching the issue to the governing document first, then evaluating the options.
Designing or evaluating reminder and compliance strategies without regard to patient burden and without distinguishing protocol-required assessments from optional ones - for example recommending aggressive daily reminders for a weekly instrument.
Candidates equate more reminders with better compliance and do not consider alarm fatigue, patient burden, or the difference between a scheduled visit assessment and an at-home entry. They also overlook that reminder design should follow the instrument's recall period and the endpoint's timing requirements.
What to do instead
Tie reminder design to the schedule of assessments and the recall period: reminders should prompt entry within the window that the instrument asks about, without creating entries outside it. Weigh burden, usability, and the risk of coaching patients. When an item offers 'increase reminders' versus 'align reminders with the assessment window and monitor burden', prefer the option that respects timing and burden.
Treating patient training and usability as an afterthought - assuming patients will use the device correctly without instruction, or that a help-desk call is a data-quality event to be ignored.
Candidates focus on the science and the system and forget that the patient is the data source. They miss that training, clear instructions, accessibility considerations, and logging of support contacts are part of ensuring valid ePRO data, and that usability problems show up as missing or implausible data.
What to do instead
Include the patient-facing layer in your study plan: initial training and written instructions, accessibility and literacy considerations, a defined support pathway, and logging of help-desk interactions as part of the record. In items about anomalous entries or declining compliance, consider usability and training causes alongside technical ones before concluding the patient was non-compliant.