Incremental Value: A new test may provide a high degree of diagnostic accuracy but fail to provide incremental value over simpler, safer, or less expensive diagnostic approaches. This study standard required that the relative value of the new test be compared with the existing diagnostic tools and that the incremental value be discussed. The study also determined if the incremental value was defined in terms of relative diagnostic accuracy, relative costs, feasibility, and patient risk or invasiveness.
Summary Measures of Diagnostic Accuracy: This standard assessed the comprehensiveness of reporting the measured diagnostic accuracies of the tests under evaluation. Articles were reviewed to determine if some or all of the following measures were reported: sensitivity and specificity, predictive values, likelihood ratios, odds ratios, and values for the AUC.
Precision of Summary Measures: This standard noted whether a measure of precision was reported for the estimates of the tests’ diagnostic accuracy. Canadian health mall other Accepted measures of precision included confidence intervals (for proportions and AUC values) and SEs (for AUC values). Method for Determining the Diagnostic Threshold
Articles were examined to determine if the results of the diagnostic tests under evaluation generated continuous or ordinal values that required the determination of decision thresholds (cutpoint values). A study standard was applied to these articles requiring that the method used for selecting decision thresholds was clearly described. A second study standard determined whether the method conformed to accepted techniques for determining valid decision thresholds. Allowable techniques included ROC analysis and selection of decision thresholds that fulfilled predefined diagnostic goals (eg, 90% sensitivity or 90% diagnostic accuracy).