Diagnostic and Screening Test (ppt).pptVIP

* This example illustrates the interplay between sensitivity and specificity. If we lower the threshold for diagnosing diabetes, we become more sensitive because fewer mild diabetics escape diagnosis by having a false negative test. The disadvantage of this is that some people without diabetes will be above the threshold and receive a false positive test. Thus, a greater percentage of those with a positive test will be false positives. Conversely, if we raise the threshold, we can be very certain that few healthy persons will be above it and have a false positive test. Thus, virtually all healthy persons will have a true negative test result. Also, virtually all of those who receive a positive test will really be diabetics. Thus, specificity is high. The disadvantage of raising the threshold is that we will miss mild diabetics, who will receive a false negative result. The bottom line is, when setting the threshold cutoff, we have to decide whether we want to have a highly sensitive test with the advantage of catching all the diabetics, but with the disadvantage of including many false positives or a highly specific test with the advantage that virtually all of our positives are true positives, but with the disadvantage of missing mild cases . Have the students discuss under which conditions you want to maximize sensitivity disease cheap to treat, few side effects of treatment, missing disease leads to bad consequences—you would not want to miss any cases and are willing to treat some false positives and when you would want to maximize specificity treatment is poorly effective or carries severe side effects—you would not want to treat a false positive. * * * * * In this case, only 69.2% of the positive tests are true positives. While imperfect, this is much better than what we started with! Without the test, we were dealing with a population in which 20% had the disease. Now, limiting ourselves to the portion of the population that had a positive test,