Table 3.
The relationship between direct, practical measures of a non-invasive screening test’s accuracy, how each measure informs assessment of diagnostic accuracy (sensitivity and specificity) and the consequence of the result for a screening programme
| Test result | Diagnostic verification* | Result of diagnostic verification | Related accuracy characteristic | Programme consequence |
| Positive | Yes | True, hence TP (TP1, TP2)† | Relative sensitivity; TP1/TP2 | Detection of neoplasia |
| Positive predictive value; TP1/(TP1+FP1) TP2/(TP2+FP2) |
Efficiency of detection | |||
| False, hence FP (FP1, FP2) | Relative FP rates; FP1/FP2 | Burden associated with detection | ||
| Negative | No | |||
| Test result | Diagnostic verification* | Result of diagnostic verification | Related accuracy characteristic | Programme consequence |
| Positive | Yes | True, hence TP | Sensitivity; TP/(TP+FN) |
Detection of neoplasia |
| Positive predictive value; TP/(TP+FP) |
Efficiency of detection | |||
| False, hence FP | Burden associated with detection | |||
| Negative | Yes | True, hence TN | Specificity; TN/(TN+FP) | Accuracy of detection |
| Negative predictive value; TN/(TN+FN) |
Exclusion of neoplasia | |||
| False, hence FN | Miss rate (missed lesions) |
The upper section shows relationships when paired testing is undertaken and only test-positive cases undergo colonoscopy while the lower section refers to the situation when all cases undergo colonoscopy.
*Verified by colonoscopy.
†TP1,TP2—TP by tests 1 and 2; FP1,FP2—FP by tests 1 and 2.
FN, false negative; FP, false positive; TN, true negative; TP, true positive.