A diary endpoint is only as good as the entries behind it. This guide defines compliance, shows how to measure it in a way that survives scrutiny, and walks through the design decisions that decide whether participants complete the diary or quietly stop.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Expected entries
1,260
Completed in window
1,071
Late
58
Missed
131
Drop after week 6
The short version
Definitions
People use "ePRO compliance" for several different things, which is why compliance figures from two studies rarely compare. Pick one definition, write it into the protocol or data management plan and report the same three numbers throughout the study.
The first is on-time completion: entries submitted inside the window the protocol allows, divided by entries expected. The second is late completion: entries submitted after the window but before the visit, which may be usable for some endpoints and not for others. The third is missed: expected entries with no submission at all. Reporting only a single blended percentage hides the difference between a participant who is a day behind and one who has stopped.
A 12-week study asks 15 participants for a daily evening diary: 15 x 84 = 1,260 expected entries. If 1,071 arrive inside the window, 58 arrive late and 131 never arrive, on-time compliance is 85 percent, late is 4.6 percent and missed is 10.4 percent. The overall figure looks healthy, yet the weekly view in the chart above shows the study losing ground, and that is the number that tells you when to act.
Endpoints are usually calculated per participant per period, for example a weekly mean pain score that needs at least four of seven daily entries. Define that rule in the statistical analysis plan. A study with 85 percent average compliance can still lose many participants-weeks if the missing entries cluster in the same people.
At-risk participants
Why it matters
The classic evidence is the work of Stone and colleagues, published in the BMJ in 2002 with a fuller report in Controlled Clinical Trials. They gave chronic pain patients paper diaries fitted with sensors that logged when the binder was opened, alongside an electronic diary that time-stamped each entry. Reported completion on paper was far higher than the records of actual use suggested, with entries evidently completed in bulk or from memory, while the electronic diary recorded high completion on its own time-stamps. The lesson is not that patients are dishonest. It is that a paper diary cannot show when an entry was made, and that a recall-based entry is not the same measurement as one made at the time.
That is the reason endpoint diaries in regulated trials lean towards electronic capture with a visible timestamp, and towards completion windows that stop participants filling in a week at once. See how to run patient diaries in clinical trials for the full set-up and how to capture patient-reported outcomes digitally for the data side.
Root causes
Most gaps trace to a handful of avoidable causes. Look at the pattern in your own data before you reach for more reminders.
| What you see | Likely cause | Design response |
|---|---|---|
| Low completion from day 1 for a few participants | Never got the link, or a barrier at first use | Test the link on the participant’s own phone at the visit; have the site confirm the first entry the same day |
| Steady decay after week 4 to 6 | Fatigue, the diary feels long | Shorten with skip logic so a good day takes under a minute; keep the full set for the days that matter |
| Entries cluster just before visits | Back-filling from memory | Use a completion window that closes, and report late entries separately |
| Evening diary missed on weekends | Routine changes | Set the due time to fit the recall period and the participant’s routine; consider a morning option for sleep items |
| Whole days missing for one participant | Travel, illness, new phone, lost interest | Flag the participant early and call; ask what is in the way |
| Dropped entries mid-form | Form too long or confusing | Cut items that do not feed an endpoint; pilot with real patients |
Pattern-reading like this only works if compliance is visible by participant and by week while the study is running, not after database lock.
Design levers in Capture
Capture sends reminders by email or SMS with a secure link that opens the questionnaire in the phone browser, so there is no app to install or password to forget. Visit and task windows decide what happens when an entry is late. Your team sees completion by participant and site while the study runs.
Enrol yourself as a test participant in the free sandbox and complete a diary from your phone. It is the fastest way to find out whether your diary is too long.
Before the first patient
Do this at protocol and build stage, when changes are cheap.
State the minimum entries per period that make the endpoint usable, and how missing entries are treated, in the statistical analysis plan.
Today-style pain scores in the evening, sleep in the morning, weekly summaries on one fixed day.
Remove any item that does not feed an endpoint or a safety decision. Use skip logic for the rest.
Hard windows for endpoint diaries, softer for supportive questionnaires. Say in the consent form what happens when an entry is missed.
Complete it yourself in the sandbox, then pilot it with two or three people who resemble your participants.
One named person reviews at-risk participants every week and calls early.
Monitoring
By week, so decay is visible early.
A site call within a working day, not a longer reminder ladder.
Check the link, the phone number or email, and the first-use experience.
A sign of back-filling; consider tightening the window.
A site with much lower completion may need a different onboarding routine.
Record why entries were missed so the analysis can separate illness from disengagement.
There is no universal threshold. The right figure depends on the endpoint and on how much missing data the analysis can tolerate. Set the minimum valid entries per period in the protocol and statistical analysis plan, and treat any benchmark you see quoted as a starting point to check, not a rule.
Entries completed inside the allowed window divided by entries expected, reported alongside late and missed entries. Calculate it by participant and by week as well as for the whole study.
Electronic entries carry a timestamp, so you can see when the entry was made. Research on instrumented paper diaries found large gaps between reported and actual completion, and recall-based entries are a different measurement from entries made at the time.
They help with forgotten entries, not with a diary that is too long, a window that does not suit the participant or a first-use barrier. Look at the pattern of missing data first, then decide whether reminders, design or a site call is the right response.
No. Capture sends an email or SMS link that opens the questionnaire in the phone browser.
Yes. Completion and overdue tasks are visible by participant and site in the study dashboard, so you can call people before gaps grow.
That is a statistical question for your protocol and analysis plan: the rules for valid weeks, imputation and sensitivity analyses. Capture records which entries were on time, late or missed so the rule can be applied; it does not decide the method.
Yes. The free sandbox has every feature, with no credit card. You pay only when you go live with real participants.
Keep exploring
ePRO reminders by SMS and email
The reminder ladder and quiet hours.
Patient diary software
Recurring diaries in a trial.
How to run patient diaries
Set-up from protocol to review.
BYOD ePRO, no app download
Phone-browser questionnaires.
Visit window calculator
Plan visit and task windows.
Sleep diary template
A morning diary to copy.
Real-time trial dashboards
Compliance at a glance.
Free sandbox with every feature. No credit card, and you pay only when you go live with real participants.