Guide, published
Out-of-window visits: how to document deviations
Preserve actual visit dates, assess the protocol issue and record the review without disguising a late visit or inventing a universal reporting rule.
In this guide
- Check the actual protocol-defined window
- Preserve what happened
- Describe the issue without a universal severity label
- Write a reconstructable deviation record
- Reporting and corrective action are study-specific
- Use the electronic workflow without replacing judgment
- Review one event from discovery to follow-up
Check the actual protocol-defined window
Start with the schedule of assessments, its footnotes and the applicable amendment. Confirm the anchor event, Day 1 or Day 0 convention and the allowed early and late dates for the particular assessment. A calendar appointment is not the protocol window. Some procedures have different timing requirements from the visit as a whole. Establish which requirement was missed before deciding how to describe the event.
The visit-window calculator can help reproduce calendar-day arithmetic. It does not interpret protocol language or decide whether an exception is important. Keep the inputs with the result: anchor, convention, target day and allowances. If the protocol is unclear, obtain the sponsor’s clarification and retain it. Do not silently move the anchor or apply a convenient default such as plus or minus three days to every study.
Preserve what happened
Record the actual date the participant attended and the actual dates of the affected procedures. A late visit should not disappear from the schedule, and its date should not be changed to fall within the window. Retain source evidence describing what was performed, omitted or deferred. A planned date and an actual date answer different questions; overwriting one with the other can hide the event that needs review.
Illustrative example: a Day 15 visit has a permitted range from 12 through 18 January, but the participant attends on 20 January. The record should show 20 January as the actual visit date and retain the original target and window. Staff review the two-day departure from the closing date and the affected assessments. This is a sample schedule, not a recommended classification or proof that every such event has the same consequence.
Describe the issue without a universal severity label
ICH E6(R3) supports protocol compliance, participant protection and reliable results. The significance of a missed window depends on the protocol and what was affected. An out-of-window visit is not automatically an important deviation in every study. Review the implications for safety, rights and data reliability, and involve the investigator and sponsor as the study procedures require.
Identify the assessment, expected timing, actual timing and known reason. Separate facts from an interpretation that is still under review. Avoid using major or minor as an unexplained label unless the study’s procedure defines it. The December 2024 FDA protocol-deviations document was issued as draft guidance marked not for implementation; do not describe its proposed classifications as a binding rule. Check the current guidance status when establishing procedures.
Write a reconstructable deviation record
A useful record identifies study, participant code, protocol version, expected event, actual event, discovery date, responsible reviewer and supporting evidence. Describe the reason if known and what was done immediately. A statement such as visit late is not enough to reconstruct whether the affected procedure was still performed or whether staff assessed participant safety. Keep the record linked to the relevant visit rather than relying on a separate email thread.
If information is learned later, add it using the controlled change procedure with the real date and reason. Do not backdate the explanation or erase the first description. A note to file can explain context, but it does not replace a deviation record, source correction or required report. The investigator’s review and the relevant reporting decision should be retained in the appropriate study workflow.
Reporting and corrective action are study-specific
Determine reporting routes and timelines from the protocol, sponsor requirements, IRB procedures and applicable regulations. There is no single deadline or automatic FDA report for every late appointment. An urgent participant-safety concern needs the study’s safety response, not just a calendar flag. Document who assessed the event and why a report was or was not required under the governing procedure.
Corrective action should address the reason rather than simply add a note saying staff retrained. If a visit was booked at the edge of the window, review appointment practices. If an anchor was wrong, review how it was selected and whether other participants are affected. If the participant could not attend, record the circumstances without blame and assess the protocol impact. Check whether the action was implemented and whether recurring issues need broader review.
Use the electronic workflow without replacing judgment
Kyros Trials shows visit targets and allowed dates anchored to Day 1 and flags out-of-window work. Staff preserve the actual visit and follow the deviation process. The software does not decide the event’s regulatory importance, approve an exception or determine the investigator’s reporting obligation. Review sample records and access permissions before relying on the workflow for a study.
Use the CTMS page, window guide and note template to prepare your team’s review. Retain the original window, actual procedure record, explanation and follow-up together. Before closing the issue, confirm that a reviewer can reproduce the chronology and find the decision. This guide provides a documentation framework, not a substitute for protocol interpretation, investigator oversight or IRB and sponsor requirements.
Review one event from discovery to follow-up
Build a sample chronology before closing a deviation. Record the planned window, actual date, discovery date, source evidence and the person who assessed the effect on the study. Confirm that the chronology distinguishes a missed appointment from a procedure that occurred but was entered late. Those events can have different consequences, and a short calendar label cannot make the distinction on its own.
For the sample late visit, ask the coordinator to identify affected assessments and the investigator to review the clinical implications. Ask the study’s reporting owner to document the governing procedure and the resulting decision. Retain follow-up actions and their completion dates. Do not invent a sponsor approval or use a note to imply that the protocol was followed when it was not. The goal is a record that another reviewer can reconstruct without changing the actual event.