Data quality and completeness in a web stroke registry as the basis for data and process mining

Electronic health records often show missing values and errors jeopardizing their effective exploitation. We illustrate the re-engineering process needed to improve the data quality of a web-based, multicentric stroke registry by proposing a knowledge-based data entry support able to help users to homogeneously interpret data items, and to prevent and detect treacherous errors. The re-engineering also improves stroke units coordination and networking, through ancillary tools for monitoring patient enrollments, calculating stroke care indicators, analyzing compliance with clinical practice guidelines, and entering stroke units profiles. Finally we report on some statistics, such as calculation of indicators for assessing the quality of stroke care, data mining for knowledge discovery, and process mining for comparing different processes of care delivery. The most important results of the re-engineering are an improved user experience with data entry, and a definitely better data quality that guarantees the reliability of data analyses.