gms | German Medical Science

68. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

17.09. - 21.09.23, Heilbronn

The estimand framework (theoretical and empirical) for diagnostic accuracy studies

Meeting Abstract

Suche in Medline nach

  • Alexander Fierenz - Department of Medical Biometry, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
  • Britta Rackow - Universität Magdeburg, Magdeburg, Germany; Department of Medical Biometry, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
  • Antonia Zapf - Department of Medical Biometry, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 68. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS). Heilbronn, 17.-21.09.2023. Düsseldorf: German Medical Science GMS Publishing House; 2023. DocAbstr. 61

doi: 10.3205/23gmds159, urn:nbn:de:0183-23gmds1597

Veröffentlicht: 15. September 2023

© 2023 Fierenz et al.
Dieser Artikel ist ein Open-Access-Artikel und steht unter den Lizenzbedingungen der Creative Commons Attribution 4.0 License (Namensnennung). Lizenz-Angaben siehe http://creativecommons.org/licenses/by/4.0/.


Gliederung

Text

The aim of a diagnostic test is to evaluate the health condition of an individual based on clinical information such as symptoms, laboratory values or physical examinations. In confirmatory diagnostic accuracy studies the diagnostic test is evaluated in individuals, where the true disease state is unknown at study inclusion. This allows to assess the performance of this test in distinguishing between the two disease states.

To transfer the trial result of the study to the real world, it is important to define the test accuracy precisely. This means, it has to be stated, which test has to be conducted for which population to reveal which condition. Moreover, different events could occur like for example rejection of the test or application of the test differently than instructed. This can lead to a non-existent or influenced test result.

The trial objective is translated into the clinical question of interest by defining an appropriate estimand. Our framework differentiates between the theoretical and empirical estimand. The theoretical estimand defines the effect in a theoretical setting based on unobservable data. The empirical estimand, however, uses observable realizations of real world data from a study to quantify this effect. Based on assumptions, the empirical estimand should align with the theoretical estimand to get external valid results. For example, the theoretical estimand defines the target population for the diagnostic test. The study population in the empirical estimand should be as close as possible to this target population.

In this talk the attributes of the theoretical and empirical estimand are defined as well as their linkage. Based on examples, the disparity between various estimands in diagnostic accuracy studies will be visualized.

It turns out, that it is important to define the study objective in the estimand in order to estimate an accuracy that is appropriate to the clinical question of interest. With the wrong choice of the estimand or a bad connection between the theoretical and empirical estimand the validity of the study results with respect to the trial objective gets lost.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.