Reporting results from studies evaluating diagnostic tests

doi:10.1016/S0196-4399(02)80013-7

Clinical Microbiology Newsletter

Volume 24, Issue 8, 15 April 2002, Pages 60-63

https://doi.org/10.1016/S0196-4399(02)80013-7 Get rights and content

Abstract

Evaluating a new diagnostic test requires careful planning. It involves choosing the appropriate comparative procedure, patients, specimens, and individuals performing the tests. The type of study design used to evaluate a new test has a direct impact on how the study results can be reported. This article describes some statistically appropriate and inappropriate practices for reporting results from different studies evaluating qualitative diagnostic tests. Special attention is given to describing a practice called discrepant resolution and its associated problems.

References (20)

A. Hadgu
Discrepant analysis: a biased and an unscientific method for estimating test sensitivity and specificity
J. Clin. Epidemiol.
(1999)
W.C. Miller
Bias in discrepant analysis: when two wrongs don't make a right
J. Clin. Epidemiol.
(1998)
M. Staquet
Methodology for the assessment of new dichotomous diagnostic tests
J. Chron. Dis.
(1981)
S.D. Walter et al.
Estimation of test error rates, disease prevalence and relative risk from misclassified data: a review
Clin. Epidemiol.
(1988)
T.A. Lang et al.
(1997)
A.J. Saah et al.
“Sensitivity” and “specificity” reconsidered: the meaning of these terms in analytical and diagnostic settings
Ann. Intern. Med.
(1997)
C.G. Begg
Biases in the assessment of diagnostic tests
Stat. Med.
(1987)
M.M. Shoukri
I. Bross
Misclassification in 22 tables
Biometrics
(1954)
J.J. Gart et al.
Comparison of a screening test and a reference test in epidemiologic studies. II. A probabilistic model for the comparison of diagnostic tests
Am. J. Epidemiol.
(1966)

There are more references available in the full text version of this article.

Cited by (2)

A PCR method for the identification of methicillin-resistant Staphylococcus aureus (MRSA) from screening swabs
2004, Pathology
The aim of this study was to assess the discriminatory power and potential turn around time (TAT) of a PCR-based method for the detection of methicillin-resistant Staphylococcus aureus (MRSA) from screening swabs.
Screening swabs were examined using the current laboratory protocol of direct culture on mannitol salt agar supplemented with oxacillin (MSAO-direct). The PCR method involved pre-incubation in broth for 4 hours followed by a multiplex PCR with primers directed to mecA and nuc genes of MRSA. The reference standard was determined by pre-incubation in broth for 4 hours followed by culture on MSAO (MSAO-broth).
A total of 256 swabs was analysed. The rates of detection of MRSA using MSAO-direct, MSAO-broth and PCR were 10.2, 13.3 and 10.2%, respectively. For PCR, the sensitivity, specificity, positive predictive value and negative predictive values were 66.7% (95%CI 51.9-83.3%), 98.6% (95%CI 97.1-100%), 84.6% (95%CI 76.2-100%) and 95.2% (95%CI 92.4-98.0%), respectively, and these results were almost identical to those obtained from MSAO-direct. The agreement between MSAO-direct and PCR was 61.5% (95%CI 42.8-80.2%) for positive results, 95.6% (95%CI 93.0-98.2%) for negative results and overall was 92.2% (95%CI 88.9-95.5%).
(1) The discriminatory power of PCR and MSAO-direct is similar but the level of agreement, especially for true positive results, is low. (2) The potential TAT for the PCR method provides a marked advantage over conventional methods. (3) Further modifications to the PCR method such as increased broth incubation time, use of selective broth and adaptation to real-time PCR may lead to improvement in sensitivity and TAT.
Diagnostic testing
2017, Veterinary Epidemiology: Fourth Edition

^☆: The views expressed in this article are those of the author and do not necessarily reflect those of the Food and Drug Administration.

View full text

Reporting results from studies evaluating diagnostic tests☆

Abstract

J. Clin. Epidemiol.

J. Clin. Epidemiol.

J. Chron. Dis.

Clin. Epidemiol.

“Sensitivity” and “specificity” reconsidered: the meaning of these terms in analytical and diagnostic settings

Ann. Intern. Med.

Biases in the assessment of diagnostic tests

Stat. Med.

Misclassification in 22 tables

Biometrics

Comparison of a screening test and a reference test in epidemiologic studies. II. A probabilistic model for the comparison of diagnostic tests

Am. J. Epidemiol.