Diagnostic Tests

Tests with Two Results, Positive versus Negative

A test result is said to be “positive” if it shows a particular finding to be present and “negative” if the finding is absent. Note that a positive test result suggests that a patient has the disease in question—which is usually not a positive thing for the patient—and vice versa.

Most diagnostic information is not perfect but rather subject to some degree of error. A positive test result may be

• true positive (TP)—the test result indicates disease, and the patient has the disease, or
• false positive (FP)—the test result indicates disease, but the patient does not have the disease.

A negative test result may be

• true negative (TN)—the test result indicates no disease, and the patient has no disease, or
• false negative (FN)—the test result indicates no disease, but the patient has the disease.

Whether a patient has the disease or not is determined by the “truth” as established by a reference (gold) standard test, which is generally an invasive and/or expensive test and one that many patients would like to avoid. The occurrence of false-positive (FP) and false-negative (FN) test results implies that medical professionals need to be careful in interpreting diagnostic test information to minimize the impact of such errors.

Diagnostic performance (also referred to as accuracy or validity) of a test is its correspondence with the underlying truth and is expressed using the test's characteristics, sensitivity, and specificity. Alternatively, the diagnostic test performance may be characterized with true- and false-positive ratios, which is particularly convenient when a test has more than two possible results. Sensitivity and specificity describe how often the test is correct in the diseased and nondiseased groups, respectively. True- and false-positive ratios describe how often the test yields a positive result in the diseased and nondiseased groups, respectively.

Sensitivity, or true-positive ratio (TPR), is the probability of a positive test result given that the disease is present, denoted by p(T+|D+). Specificity, or true-negative ratio (TNR), is the probability of a negative test result given that the disease is absent, denoted by p(T-|D-). The false-negative ratio (FNR) is the complement of sensitivity, that is, 1.0 – TPR, and is the proportion of patients with disease who have a negative test result, denoted by p(T-|D+). The false-positive ratio (FPR) is the complement of specificity, that is, 1.0 – TNR, and is the proportion of patients without disease who have a positive test result, denoted by p(T+|D-).

...