Harmonized data quality reporting made accessible
Abstract Data quality reporting is frequently non-transparent in current research practice. This impairs reproducible research. One of the many contributing factors to this situation is uncertainty (1) on how to set up data appropriately for data quality assessments, and (2) on how to efficiently conduct such assessments. This talk will first provide a definition of data quality and illustrate the impossibility of any static approach to assess data quality. Second, an overview will be given on selected data quality frameworks of relevance in the medical sciences to illustrate the complexity of comprehensive data quality assessments. Finally, tools (R, Stata) will be presented that enable efficient data quality assessments to promote harmonized data quality reporting.
Dr. Carsten Oliver Schmidt is Professor at the Institute of Community Medicine, University Medicine of Greifswald. He is deputy head of the Department of SHIP/ Clinical-Epidemiological Research and leads the functional division “Quality in the Health Sciences”. He is PI of several methods and research quality related projects, and speaker of the methods section of the German Society for Epidemiology. He is responsible for quality management procedures in population based observational health studies, has expertise in data linkage of primary and secondary data including registries, claims data, hospitals, and in the development of research infrastructures. Research topics include research methods, the study of subclinical and clinical disorders, incidental findings.