Database lock represents a critical milestone in clinical trial conduct, signifying that all protocol-required data have been collected and entered, all queries have been resolved, all data cleaning activities have been completed, and the data are ready for statistical analysis. After lock, the database becomes read-only and no further modifications are permitted without formal unlock procedures, ensuring that the data analyzed and reported are stable and verifiable. The locked database serves as the foundation for all efficacy and safety analyses and ultimately supports regulatory submissions.

The activities leading to database lock follow a structured process designed to ensure data quality and completeness. Source document verification confirms that entered data match original records. Medical coding ensures that adverse events and concomitant medications are coded using standardized terminologies. Data cleaning resolves all outstanding queries and verifies logical consistency across data fields. Serious adverse event reconciliation confirms that all SAEs are correctly captured in both clinical and safety databases. External data from central laboratories, interactive response technology, and other sources are integrated and reconciled with site-entered data.

Database lock typically proceeds through defined phases with quality gates ensuring readiness at each step. Soft lock may restrict entry of new data while allowing resolution of existing queries. Pre-lock review by clinical, data management, and biostatistics teams verifies completeness and quality. Hard lock formally closes the database and may involve electronic signatures by designated personnel attesting to data readiness. Documentation of the lock process, including verification of completion criteria and authorization signatures, forms part of the essential trial documents. Any subsequent need to modify locked data requires formal unlock procedures with documented justification and approval.