Assessing the quality of the data in a transplant registry: the European Liver Transplant Registry.

The usefulness of a clinical database depends strongly on the quality of the collected data. Data quality assurance should start with deciding in advance the uses to which the database is going to be put, developing an explicitly defined minimum data set, and setting up a user-friendly interface. The quality of the collected data can be assessed by computer validation, during which computerized range and consistency checks are based on information within the database itself. Also, note validation can be used, which implies a comparison of the original database against medical records. A number of related methodological issues are discussed in this article.