Wallach's Interpretation of Diagnostic Tests: Pathways to Arriving at a Clinical Diagnosis Read Online Free PDF Page A

false-negative results. Specificity is defined as the ability of the test to identify correctly those who do not have the disease. It is the number of subjects who have a negative test and do not have the disease divided by all subjects who do not have the disease. A test with high specificity has few false-positive results. Sensitivity and specificity are most useful when assessing a test used to screen a free-living population. These test characteristics are also interdependent (Figure 1-1 ): an increase in sensitivity is accompanied by a decrease in specificity and vice versa.

    Figure 1–1 Sensitivity, specificity, and predictive values in laboratory testing. NPV, negative predictive value; PPV, positive predictive value.
    Predictive values are important for assessing how useful a test will be in the clinical setting at the individual patient level. The positive predictive value (PPV) is the probability of disease in a patient with a positive test. Conversely, the negative predictive value (NPV) is the probability that the patient does not have disease if he has a negative test result.
    PPV and sensitivity of tests are complementary in their examination of true positives. Given that the test is positive, PPV is the probability that the disease is present, in contrast to sensitivity, which is given that the disease is present, the probability that test is positive. Likewise, NPV and specificity are complementary in their examination of true negatives. Given that the test is negative, NPV is the probability that the disease is absent. This is in contrast to specificity, which is given that the disease is absent, the probability that test is negative (see Figure 1-1 for more information). Predictive values depend on the prevalence of a disease in a population. A test with a given sensitivity and specificity can have different predictive values in different patient populations. If the test is used in a population with high disease prevalence, it will have a high PPV; the same test will have a low PPV when used in a population with low disease prevalence.
    Likelihood ratios (LRs) are another way of assessing the accuracy of a test in the clinical setting. They are also independent of disease prevalence. LR indicates how much a given diagnostic test result will raise or lower the odds of having a disease relative to the probability of the disease. Each test is characterized by two LRs: positive LR (PLR) and negative LR (NLR). PLR tells us the odds of the disease if the test result is positive, and NLR tells the odds of disease if the test result is negative.

    An LR >1 increases the odds that the person has the target disease, and the higher the LR, the greater this increase in odds. Conversely, an LR ratio <1 diminishes the odds that the patient has the target disease.
    RECEIVER OPERATING CHARACTERISTIC (ROC) CURVES
    ROC curves allow one to identify the cutoff value that minimizes both false positives and false negatives. An ROC curve plots sensitivity on the y axis and 1 − specificity on the x axis. Applying a variety of cutoff values to the same reference population allows one to generate the curve. The perfect test would have a cutoff value that allowed an exact split of diseased and nondiseased populations (i.e., a cutoff that gave both 100% sensitivity and 100% specificity). It would plot as a right angle with the fulcrum in the far upper left corner (x = 0, y = 1). This case, however, is very rare. For most cases, as one moves from the left to right on the ROC curve, the sensitivity increases and the specificity decreases.
    Calculation of the area under the ROC curve allows comparison of different tests. A perfect test has an area under the curve (AUC) equal to 1. Therefore, the closer the AUC is to 1, the better the test. Similarly, if one wants to know the cutoff value for a test that minimizes both false positives and false negatives (and hence maximizes both sensitivity and specificity), one would select the point