Information about participation changes should be regularly reviewed by trial oversight groups during each trial.
The frequency of review will depend on the trial and its risks. However, it should be frequent enough to allow any corrective actions to be taken (and have time to be effective) in response to any issues raised through the review.
Reporting participation change data for monitoring purposes
The terminology used should convey the complexity in participation changes (see principle O1 – participation can stop, reduce or change).
Presented data should separate different types of participation change. This might include participants stopping intervention only, stopping further follow-up but allowing use of data from routine follow-up, stopping all data collection, or having lost contact with trial staff. The exact categories may depend on the trial. They may also align with the ‘tiers’ of participation (see Protocol Development).
Prioritise data about issues that could affect completeness or validity of the primary outcome measure(s).
Reviewed data should summarise information on reasons for participation changes, when available (see Data Collection about Participation Changes), including where the reason is unknown.
In any trials with more than one intervention group, data should also be presented and reviewed by treatment group. It may be more appropriate for this information to be restricted to the Independent Data Monitoring Committee (particularly for randomised trials), rather than being accessed by members of the trial team.
Data should be reviewed by trial site and based on any other important subdivisions for a given trial, e.g. by country in international trials.
Information about baseline characteristics should be included with any of these statistics as required. This can show any differences in characteristics between groups of participants who have missing outcome data and those who do not.
As always, data reports generated during an ongoing trial should be interpreted with some caution, as data collection and cleaning will be ongoing.
The advice above (as per the rest of this guidance) mostly focuses on participation changes. Missing data may occur for other reasons, such as participants not attending some trial visits (without fully stopping their participation) or administrative or technical reasons. Summarised information about the occurrence of missing data (with any cause) should be reviewed on a regular basis throughout the trial.