Table 1

Methodological standards for the evaluation of diagnostic tests. Standards 1, 2, and 7 relate to the relevance of the results to particular settings, whereas standards 3, 4, and 6 are primarily about the validity of the results. Standard 5 relates to the need for careful reporting of the results

(1)  Specification of spectrum composition This standard requires at least three of the following four descriptors to be reported for the study population: the age and sex distribution, the presenting clinical symptoms and/or disease stage of the populations studied, and the eligibility criteria for the subjects included.
(2)  Analysis of pertinent subgroups This standard requires the evaluation to cite the indices of accuracy for any pertinent demographic or clinical subgroup of the population.
(3)  Avoidance of work-up (verification) bias This standard requires an evaluation to ensure all subjects receive both diagnostic testing and gold standard verification.
(4)  Avoidance of review biasThis standard requires an evaluation to make a clear statement about the independence in interpreting both the test and the gold standard procedure.
(5)  Presentation of precision of results for test accuracy This standard requires an evaluation to report the 95% CI or standard error associated with the indices of diagnostic accuracy
(6)  Presentation of indeterminate test results This standard requires an evaluation to state the number of indeterminate results and whether or not these results had been included or excluded when the indexes of accuracy were calculated.
(7)  Presentation of test reproducibility This standard requires that the reproducibility of a test should be reported, or that the report should cite other sources of this information.
  • Jaeschke et al 2 3 and Reid et al.4