Data extraction from a data warehouse at a German university hospital – data quality and current challenges

Treatment of chronic wounds is a topic with high research potential, however, routine data are hardly standardized and cannot be accessed easily. The introduction of data integration centers within the Medical Informatics Initiative (MII) in Germany aims at making routine care data available for research. In this study, we seek to investigate the degree of structuredness and the quality of clinical data for clinical research on chronic wounds. These data were extracted in csv format from the data warehouse in the data integration center at the University Hospital in Erlangen. We found that data retrieved for wound research were heterogeneous in format (free text vs. structured data) and quality (missing values, inconsistencies, lack of longitudinal data, missing context) and were mainly based on billing relevant data. Basic epidemiological and clinical questions might be answered, but necessary data are missing for deeper analyses of treatment outcomes. A more clinically driven and standardized documentation would improve the availability of data relevant for research on the courses of treatments in the field of wound care and beyond.