Project Details
Projekt
Standards and tools for data monitoring in observational studies
Applicants
Professor Dr. Martin Dugas; Professorin Dr. Iris Pigeot; Professor Dr. Wilhelm Sauerbrei; Professor Dr. Carsten Oliver Schmidt; Privatdozent Dr. Henning Silber
Subject Area
Medical Informatics and Medical Bioinformatics
Epidemiology and Medical Biometry/Statistics
Epidemiology and Medical Biometry/Statistics
Term
since 2016
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 315057723
The first funding period has stimulated an extensive exchange between representatives of major German cohort studies on the conduct and optimization of data quality assessments. The ground has been laid for the first data quality assessment approach in epidemiology linking a data quality framework with generic statistical implementations. This comprised (1) an assessment of the German TMF guideline on data quality, which resulted in a revised data quality concept; (2) statistical implementations in R and Stata; (3) the extension of a web application for data quality assessments. Several public workshops were conducted and a web portal was established to disseminate project results. Work packages in the second funding period target expanded analysis tools, cross-disciplinary standards and guidance materials to foster the sustainable and widespread use of our developments on harmonized data quality analyses in cohort studies and observational health research. The first objective improves the scope and methodology of data quality assessments. We will improve transdisciplinary exchange where we exploit the fact that epidemiology and the social sciences share in common many methods. GESIS will contribute their expertise to reveal important yet uncovered issues in our data quality concept such as adverse response behaviours. Vice versa, no comparable data quality framework exists in the social sciences. Our data quality concept may be of substantial use for observational studies in this field. Second, we will derive methods for the automated grading of data quality issues with a focus on observer-, device- and centre-effects as well as time-trends. The second objective targets the FAIRness – findability, accessibility, interoperability, and reusability - of data quality assessments. Our first goal within this objective is to overcome the limited transferability of data quality-related metadata between studies. Harmonized metadata standards will be developed in cooperation with the worlds’ largest repository on medical forms (MDM) along import and export functionalities to increase their reusability. The second goal is to ease the application of our tools. Since many persons responsible for data quality assessments are non-statisticians, interactive front-ends will enable report generation and analysis without programming skills. The third goal is to set up an e-learning environment with an open online course to teach and train digital skills in the field of data quality assessments. Our fourth goal is to improve the visibility of results from data quality assessments in scientific papers by developing a reporting guideline in cooperation with the STRATOS initiative and the EQUATOR network. Our integration in various national and international networks will increase the outreach of our project results.
DFG Programme
Research Grants
