REDCap projects usually begin in development mode. Development mode allows the team to create and modify instruments freely while the database is being built and tested. During this phase, fields may be added, deleted, reordered, renamed, or revised. Test records may be created to check workflow and logic. Development mode is appropriate before real study data are collected.
Production mode is used once the database is ready for real data collection. Moving to production signals that the project has passed design review and testing. In production, REDCap applies more controlled processes for structural changes. This protects collected data from accidental corruption. For example, changing a variable name after data have been collected can affect exports and analysis scripts. Deleting a field can risk data loss. Modifying coded options can change interpretation.
The decision to move to production should be deliberate. Before production, the study team should confirm that instruments match approved CRFs, field labels are clear, validation rules work, branching logic behaves correctly, user roles are configured, DAGs are tested if needed, reports are available, and training has been completed. Test data should be reviewed and, depending on institutional policy, deleted or clearly separated from real data before production begins.
Once in production, changes should follow a change-control process. Minor changes, such as correcting a spelling error, may be low risk. Major changes, such as altering response options, adding outcome variables, changing branching logic, or renaming variables, may require review by the data manager, investigator, statistician, sponsor, or ethics committee. The change should be documented in a version log with date, reason, approval, and impact.
Production mode does not mean the database can never change. Clinical research is dynamic, and amendments may be necessary. It means changes are controlled, documented, and evaluated for their effect on existing data and future analysis