Data Quality, Cleaning, and Reconciliation
What you must know
Data cleaning identifies and resolves errors, inconsistencies, and outliers. Reconciliation compares data across sources (EDC vs. lab, EDC vs. SAE database, EDC vs. external vendor) to ensure completeness and consistency. Understand the difference between cleaning (fixing errors) and reconciliation (matching sources).
Concepts to master
Know common cleaning activities: range checks, consistency checks, duplicate detection, and medical coding review. Understand the role of the data review meeting and the data cleaning plan. Recognize the importance of the serious adverse event (SAE) reconciliation and the lab reconciliation. Know that medical coding (MedDRA, WHO Drug) must be consistent and version-controlled.
Traps
Questions often confuse reconciliation with cleaning, or imply that reconciliation is optional. Watch for wording that treats coding as a data entry task rather than a review activity. Another trap: assuming an outlier is always an error when it may be a true value requiring source confirmation.
Rule of thumb
Reconcile before you lock: no database should be locked until EDC, SAE, and external data sources agree on key identifiers and events.