A patient is supposed to record their pain level every evening. Life gets in the way for three days, and on the fourth, they fill in all three missed entries from memory before the app locks them out.
Nothing about that moment looks like fraud it's an ordinary, human response to a missed task. But if that patient's diary entries feed a primary endpoint, those backfilled entries carry a specific, well-documented problem: the data no longer reflects what the patient actually experienced when it happened. It reflects what they remember, days later, which is a different and less reliable thing.
Recall bias is the systematic distortion that occurs when a person reports a past experience a symptom, a pain level, a quality-of-life measure from memory rather than at the moment it occurred. Memory doesn't just fade; it reconstructs. A patient asked to describe their pain over the last three days will tend to average, round, or anchor on the most recent or most memorable moment, rather than accurately representing each individual day. For a trial relying on daily or event-driven patient-reported data, that distortion doesn't stay a minor imprecision it can shift the very endpoint the study is designed to measure.
This is exactly why regulatory guidance on patient-reported outcomes puts so much weight on contemporaneous data capture recording the experience as close as possible to when it happens, rather than allowing it to be reconstructed afterward. A diary entry made three days late isn't the same data point as one made in the moment, even if the number a patient eventually writes down looks identical on the page.
A single late entry, on its own, is a minor operational issue. A pattern of late or backfilled entries across a study is something different: it's a signal that the data supporting a primary endpoint may not reflect real-time patient experience at all. If that pattern isn't visible until database lock when a monitor or data manager finally reviews timestamps against expected entry windows the study has already collected months of data whose reliability is now in question, with no way to go back and capture it correctly.
That's the real cost of not knowing about a compliance gap until it's too late to do anything about it. Catching one missed entry in week one is a reminder. Catching a pattern of missed entries in month six is a finding.
| Paper Diary | Basic ePRO App | Cloudbyz ePRO | |
|---|---|---|---|
| Timestamp integrity | None — entries can be written any time and dated freely | Recorded, but not always enforced against an entry window | Enforced entries are time-stamped at submission, in real time |
| Detects a missed or late entry | Not until someone manually reviews the paper diary | Sometimes, depending on the platform | Yes automated alerts flag non-compliance as it happens |
| Risk of backfilled entries | High | Moderate depends on whether the app locks past dates | Low real-time capture design and reminders reduce the opportunity to backfill |
| Sponsor visibility into compliance during the study | None until the diary is collected | Limited, usually periodic | Real-time dashboards showing compliance patterns as they form |
| Patient burden | High easy to forget, nothing prompts a return | Moderate | Lower automated reminders prompt patients before an entry is missed |
If more than one of these is uncomfortable to answer, the risk isn't hypothetical it's already accumulating in the data you're collecting right now.
Cloudbyz ePRO captures patient-reported data in real time across phones, tablets, and wearable devices, with entries time-stamped the moment they're submitted rather than whenever a patient gets around to it. Automated reminders and alerts prompt patients before an entry window closes, reducing the number of missed entries in the first place and when a pattern of non-compliance does start to form, real-time dashboards surface it to the study team immediately rather than waiting for a periodic review. Because ePRO shares a platform with Cloudbyz EDC, that patient-reported data flows directly into the centralized clinical dataset, so a compliance question about diary entries doesn't require reconciling two separate systems to investigate.
For Clinical Data Managers and CRAs, that turns compliance monitoring from a retrospective audit exercise into something visible throughout the study catching a pattern in week three instead of discovering it at lock.
If your team is still discovering ePRO compliance gaps at database lock instead of while they're forming, it's worth seeing what real-time visibility actually changes. Book a demo with Cloudbyz team.