Estimating predictive value of tests without having a Gold Standard: the concept of Etiologic Predictive Value (EPV)

Gunnarsson, Ronny, and Lanke, Jan (2016) Estimating predictive value of tests without having a Gold Standard: the concept of Etiologic Predictive Value (EPV). In: Abstracts from the 8th Biennial Are You Remotely Interested Conference. From: AYRI 2016: 8th Biennial Are You Remotely Interested Conference, 20-22 July 2016, Mount Isa, QLD, Australia.

[img]
Preview
PDF (Abstract Only) - Published Version
Download (64kB) | Preview
View at Publisher Website: https://www.jcu.edu.au/mount-isa-centre-...
 
8


Abstract

Background: Patients with infectious diseases often require the use of microbiologic diagnostic tests. Predictive value of tests are used to describe the usefulness of a diagnostic test in a specific setting. Sometimes an acceptable Gold standard is lacking making it difficult to evaluate the usefulness of a new diagnostic test.

Aims Of Study/project: Describe how predictive value of tests can be calculated despite the absence of an acceptable Gold Standard. Also to describe how to account for asymptomatic carriers.

Methods: Mathematical derivation shows that information from a healthy control population can, for most scenarios, be used to calculate predictive value of tests despite the absence of a Gold Standard.

Results: Rules for how the usefulness of diagnostic tests can be estimated in the absence of a Gold Standard. The new statistical method considers the influence asymptomatic carriers will have on the diagnostic process. These rules are especially suited for evaluating microbiologic diagnostic tests but can be used also in other scenarios.

Conclusions/ Recommendations: The existing Gold Standard should always be challenged when evaluating a new diagnostic test. Etiologic Predictive Value offers an alternative to comparing the new test with a conventional Gold Standard.

Item ID: 45364
Item Type: Conference Item (Abstract / Summary)
Keywords: predictive value of tests
Related URLs:
Date Deposited: 06 Sep 2016 23:54
FoR Codes: 11 MEDICAL AND HEALTH SCIENCES > 1117 Public Health and Health Services > 111717 Primary Health Care @ 100%
SEO Codes: 92 HEALTH > 9202 Health and Support Services > 920203 Diagnostic Methods @ 100%
Downloads: Total: 8
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page