Data validation

Data validation is the process of checking the information in the database for completeness, consistency and plausibility. At the TCC, we have developed a range of data validation procedures to ensure that study databases contain high quality data. We program logical checks into our data management system and perform manual checks where necessary. This enables us to detect deviations from the study protocol and to trace mistakes made when the data were entered. We discuss these findings with the investigators.

We prepare a data validation plan for each study. The principle investigator needs to approve this plan before data collection can begin. However, the data validation plan can be adapted if the data warrant this. The principle investigator has to re-approve the adapted plan before it is implemented.

If you use our system of electronic case report forms, there is a secure connection between the terminals study staff use and the database. Automatic warnings can be issued if a participant appears ineligible for inclusion in the study. In addition, members of your staff are immediately prompted to check incomplete, inconsistent or implausible data. If we perform data entry from paper case report forms, data validation is carried out after the data have been entered into the database.

When the data have been validated, we present the results to you. This enables you and your staff to correct errors and inconsistencies and to clarify other queries. At the end of the process, we summarise it in the data validation and data management reports.