Joint evaluation of accuracy parameters in diagnostic studies

• Applicant: Professor Dr. Harald Binder, since 1/2018
• Subject Area: Epidemiology and Medical Biometry/Statistics
• Term: from 2013 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 233159376

The joint evaluation of sensitivity and specificity of a diagnostic test is an old and widely accepted principle in diagnostic research. However, in most studies evaluating a binary diagnostic test sensitivity and specificity are analysed separately, not allowing any compensation between these two parameters, even if this is clinically meaningful.Specification of a region of interest, i.e. a set of pairs of values for both sensitivity and specificity which are desirable, allows to take the clinical context into account. Once such a region is specified, formal hypothesis tests or appropriate visualisations can be applied. It is the purpose of the project to continue to develop and implement corresponding optimal and feasible statistical procedures and to illustrate their use - including the elicitation of regions of interest from a patient perspective or a clinical perspective - in several case studies. Both single arm as well as comparative studies will be covered. The procedures should be collected in a software tool to allow a widespread and easy use in future. The methods and tools developed in this project will contribute to a more adequate and efficient analysis of studies on diagnostic tests. The gain in efficiency will allow a reduction in sample size and hence may speed up the process of establishing new diagnostic procedures. The general approach considered in this project is closely related to recent discussions about how to demonstrate the clinical benefit of new diagnostic methods in order to support policy makers in reimbursement decisions.

• DFG Programme: Research Grants
• International Connection: Denmark, United Kingdom
• Participating Persons: Professor Poul Flemming Hoilund-Carlsen; Professor Dr. Robert Newcombe; Professor Dr. Wolfgang A. Weber
• Ehemaliger Antragsteller: Professor Dr. Werner Vach, until 12/2017