Project Details
Projekt
Evaluation of diagnostic tests with spatially or temporally clustered data (ClusterDiag)
Applicants
Nicole Rübsamen, Ph.D.; Professorin Dr. Antonia Zapf
Subject Area
Epidemiology and Medical Biometry/Statistics
Term
since 2024
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 539658720
Diagnostic tests are the starting point of every clinical decision-making process and are thus responsible for the quality of care available to a patient. As a consequence, high-quality diagnostic studies are of great importance to be able to provide the most appropriate diagnostic test for a given person in a given diagnostic setup. However, statistical methods for diagnostic studies are not as well developed as those for therapeutic trials. One critical gap is the inclusion of repeated observations per person, leading to so-called clustered data, in the evaluation process when using the classic diagnostic accuracy measures of sensitivity, specificity, and the area under the ROC curve. The central aim of this project is to provide a comprehensive framework for planning and analysing diagnostic studies with spatially or temporally clustered data, and translate this framework into research practice. To do so, we will first review the currently available methods in the field of diagnostic studies with clustered data systematically and compare them using systematic simulation studies. We will then expand the currently developed estimand framework for diagnostic studies to the case of temporally and spatially clustered data. Based on the estimand concept, we will develop methodological approaches for valid estimates further and compare them systematically with available methods. Furthermore, methods for predictive values and for sample size planning will be developed. Finally, we will provide an implementation of the new framework as a guidance document and an R package to allow translation into research practice. All of this will be done in close collaboration with experts in the methodology and application of diagnostic tests.
DFG Programme
Research Grants
