How to manage queries in an EDC
A query is a question about a data point. Word it well, route it to the right person and close it fast, and your database lock gets shorter.
Interactive tour on demo data. Click through at your own pace, no email asked.
All guides and toursInside the demo
A 30-minute call on your protocol, your sites and the modules you need.
- Let edit checks raise the routine queries, and keep manual queries for review findings.
- Write each query as a question the site can answer from the source.
- Track open queries by site and by age every week.
- Close a query when the answer resolves it, and raise a new one when it does not.
A data query is a question the data management team, a monitor or the EDC sends to a site about a value that looks wrong, missing or inconsistent. The site checks the source, corrects the value or explains it, and the query closes. You manage queries well with fewer of them, clearer wording and one review round a week, so nothing piles up before database lock.
Automatic and manual queries
Automatic queries come from edit checks: the EDC raises them when a value breaks a rule. Manual queries come from people: a data manager reviewing listings, a monitor during source data verification, a medical reviewer reading adverse events. Both follow the same life cycle: the query opens, the site answers, and the reviewer closes it once the answer settles the question. In Datacapt the three states are open, resolved and closed.
How to run the query workflow in 6 steps
- Define the roles. Decide who can raise, answer and close a query: data manager, monitor, medical reviewer, site staff.
- Write standard wording. Keep a short list of query texts for the frequent cases, such as a missing date or a value out of range, in your data management plan.
- Word each query as a question. "Visit date is before consent date. Please check both dates in the source." The site should know what to open and what to change.
- Answer and close fast. Review answered queries within days. Close the query when the answer settles it. When it does not, reply in the thread, or close it and raise a new query on the same value.
- Follow ageing by site. Every week, list open queries by site and by age. Call the sites with queries older than 30 days.
- Clean before each milestone. Before an interim analysis or database lock, target zero open queries on critical data.
Metrics worth watching
- Queries per 100 data points. A high rate points to a confusing form or a check that fires too often.
- Median time to answer. Compare sites. A slow site needs training or more staff time.
- Repeat queries. A second query on the same value means the first question was unclear.
Queries in Datacapt
In Datacapt you raise a query on any value and follow it through to closure. The site sees the query the next time it opens the eCRF. When a value breaks an edit check, the site sees your message and Datacapt raises the query on that value. Monitors see which values still need source data verification once you switch on eCRF review. The inclusion list shows which visits each participant has completed, and the study dashboard follows data entry progress and open queries.
Keep reading
Who can answer a query?
Anyone whose role carries the permission to answer queries, most often site staff. A reviewer with the permission to close queries, often the monitor or data manager who raised it, reads the answer and closes the query.
How long should a query stay open?
Aim for an answer within a week or two. Agree the target with sites in the data management plan and follow it in the ageing report.
Can a query close without a change to the data?
Yes. When the source confirms the value, the site explains it in the answer and the reviewer closes the query. The value and the explanation both stay in the audit trail.