Premium
Methods for evaluating the performance of diagnostic tests in the absence of a gold standard: a latent class model approach
Author(s) -
Garrett Elizabeth S.,
Eaton William W.,
Zeger Scott
Publication year - 2002
Publication title -
statistics in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.996
H-Index - 183
eISSN - 1097-0258
pISSN - 0277-6715
DOI - 10.1002/sim.1105
Subject(s) - latent class model , gold standard (test) , medical diagnosis , latent variable , diagnostic test , latent variable model , statistics , diagnostic accuracy , test (biology) , class (philosophy) , computer science , econometrics , medicine , machine learning , artificial intelligence , mathematics , pediatrics , pathology , paleontology , biology
In many areas of medical research, ‘gold standard’ diagnostic tests do not exist and so evaluating the performance of standardized diagnostic criteria or algorithms is problematic. In this paper we propose an approach to evaluating the operating characteristics of diagnoses using a latent class model. By defining ‘true disease’ as our latent variable, we are able to estimate sensitivity, specificity and negative and positive predictive values of the diagnostic test. These methods are applied to diagnostic criteria for depression using Baltimore's Epidemiologic Catchment Area Study Wave 3 data. Copyright © 2002 John Wiley & Sons, Ltd.