Discover full dictionary

A query, in clinical data management, is a formal request for clarification or correction issued when a data point recorded in the case report form appears missing, inconsistent, out of range or implausible. Queries are generated automatically by edit checks programmed in the EDC system (for example a visit date before the enrolment date, or a laboratory value outside physiological limits), manually by data managers or medical reviewers during data review, or by monitors during source data verification. The site responds by confirming, correcting or explaining the entry; every step is recorded in the audit trail with user, timestamp and reason for change, as required by 21 CFR Part 11 and the EMA guideline on computerised systems.

Query management is a core process of the data management plan: it defines query text conventions, response timelines (often 5 to 10 working days), escalation, closure rules and metrics such as query rate per subject and ageing. High query volumes signal poor eCRF design, unclear protocol definitions or site training needs, and unresolved queries block database lock. Under ICH E6(R3) risk-based approaches, queries are concentrated on critical data and processes rather than on every data point.

Queries are a frequently overlooked channel for personal data leakage. Free-text query responses written by site staff sometimes include the participant's name, initials, hospital number or clinical details that go beyond the data point in question, which breaks the pseudonymisation boundary between site and sponsor. Good practice is to train site staff and data managers to write queries and responses that refer only to the subject number and the field concerned, to configure the EDC so that query text cannot be exported without review, and to include query text in periodic data reviews for identifying information. Query metadata also constitutes personal data about site staff and data managers, covered by the site-staff information notice.