An introduction to sensitivity, specificity, predictive values and likelihood ratios

Chu, Kevin (1999) An introduction to sensitivity, specificity, predictive values and likelihood ratios. Emergency Medicine Australasia, 11 3: 175-181. doi:10.1046/j.1442-2026.1999.00041.x


Author Chu, Kevin
Title An introduction to sensitivity, specificity, predictive values and likelihood ratios
Journal name Emergency Medicine Australasia   Check publisher's open access policy
ISSN 1035-6851
1442-2026
1742-6723
Publication date 1999-09
Sub-type Article (original research)
DOI 10.1046/j.1442-2026.1999.00041.x
Volume 11
Issue 3
Start page 175
End page 181
Total pages 7
Place of publication Richmond, Vic., Australia
Publisher Wiley-Blackwell
Language eng
Abstract Objective: To assist clinicians and other health-care providers to understand the terms used to describe the accuracy of diagnostic tests. Methodology: This paper reviews the calculation and interpretation of sensitivity, specificity, predictive values, receiver operating characteristic curves and likelihood ratios. Results: Sensitivity and specificity are measures of the accuracy of a diagnostic test. There is a trade-off between sensitivity and specificity that is dependent on the cut-off level chosen for a positive diagnosis. Predictive values are measures of the usefulness of a test once the test results are known. Predictive values depend on the prevalence of the disease, as well as on the sensitivity and specificity of the test. Sensitivity, specificity and predictive values are easily calculated by the construction of a two-by-two table. Multiple testing, either in parallel or in series, can alter the sensitivity, specificity and predictive values. Receiver operating characteristic curves plot the relationship between sensitivity and specificity at various cut-offs, facilitating the comparison of accuracy among tests. Likelihood ratios are an alternative measure of accuracy and have the advantage of being able to assess multiple test outcomes, rather than simply assessing dichotomized positive or negative results. Conclusion: An understanding of how the measurements that are used to describe the accuracy of diagnostic tests are calculated is essential to the interpretation of test results.
Keyword Likelihood ratios
Multiple testing
Negative predictive values
Positive predictive values
Receiver operating characteristic curves
Q-Index Code C1
Q-Index Status Provisional Code
Institutional Status Unknown

Document type: Journal Article
Sub-type: Article (original research)
Collection: School of Medicine Publications
 
Versions
Version Filter Type
Citation counts: Scopus Citation Count Cited 31 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Created: Wed, 02 Mar 2011, 15:33:42 EST by Dr Kevin Chu on behalf of Royal Brisbane Clinical School