Sequentially Determined Measures of Interobserver Agreement (Kappa) in Clinical Trials May Vary Independent of Changes in Observer Performance.

Cohen kappa biased estimator central reading clinical trials interobserver agreement repeated measures simulation

Journal

Therapeutic innovation & regulatory science
ISSN: 2168-4804
Titre abrégé: Ther Innov Regul Sci
Pays: Switzerland
ID NLM: 101597411

Informations de publication

Date de publication:
30 Sep 2019
Historique:
entrez: 2 10 2019
pubmed: 2 10 2019
medline: 2 10 2019
Statut: aheadofprint

Résumé

Cohen's kappa is a statistic that estimates interobserver agreement. It was originally introduced to help develop diagnostic tests. Interpretative readings of 2 observers, for example, of a mammogram or other imaging, were compared at a single point in time. It is known that kappa depends on the prevalence of disease and that, therefore, kappas across different settings are hard to compare. Using simulation, we examine an analogous situation, not previously described, that occurs in clinical trials where sequential measurements are obtained to evaluate disease progression or clinical improvement over time. We show that weighted kappa, used for multilevel outcomes, changes during the trial even if we keep the performance of the observer constant. Kappa and closely related measures can therefore only be used with great difficulty, if at all, in quality assurance in clinical trials.

Sections du résumé

BACKGROUND BACKGROUND
Cohen's kappa is a statistic that estimates interobserver agreement. It was originally introduced to help develop diagnostic tests. Interpretative readings of 2 observers, for example, of a mammogram or other imaging, were compared at a single point in time. It is known that kappa depends on the prevalence of disease and that, therefore, kappas across different settings are hard to compare.
METHODS METHODS
Using simulation, we examine an analogous situation, not previously described, that occurs in clinical trials where sequential measurements are obtained to evaluate disease progression or clinical improvement over time.
RESULTS RESULTS
We show that weighted kappa, used for multilevel outcomes, changes during the trial even if we keep the performance of the observer constant.
CONCLUSIONS CONCLUSIONS
Kappa and closely related measures can therefore only be used with great difficulty, if at all, in quality assurance in clinical trials.

Identifiants

pubmed: 31569962
doi: 10.1177/2168479019874059
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Pagination

2168479019874059

Auteurs

Russell Reeve (R)

Decision Sciences, IQVIA, Durham, NC, USA.

Klaus Gottlieb (K)

Immunology, Eli Lilly, Indianapolis, IN, USA.

Classifications MeSH