|
Post by Admin on Jan 17, 2016 19:58:02 GMT
A town with 1000 citizens has been plagued by a new bacterial upper respiratory infection that causes patients to experience severe shortness of breath for months. The prevalence of the infection in the town is 10%, and the incidence is 7 per 1000 person-years. A screening test for the responsible bacterium just came out, with a sensitivity of 80% and a specificity of 70%. If everyone in the town were tested with this new screening tool, how many people would test negative?
A.650 B.350 C.100 D.270 E.630
|
|
|
Post by Admin on Jan 17, 2016 20:07:38 GMT
Setting up a table to calculate an epidemiology problem is often useful:
A prevalence of 10% means that in a town of 1000 people, 100 individuals will actually have the disease and 900 will not. Sensitivity is calculated as ([TP] / [TP + FN]), indicating that 80% of the 100 individuals with the disease will actually test positive. By this reckoning, TP = 80 and FN = 20. Specificity is calculated as ([TN]) / [TN + FP]), meaning that 70% of the 900 individuals who do not have the disease will test as true negative (630). This leaves 270 people who will test as false positive. Completing the table, the total number of people who will test negative (regardless of whether they have the disease) is 20 + 630 = 650. B is not correct. 11% chose this. The figure 350 is the number of people who will test positive regardless of whether they have the disease. This is the sum of true-positive and false-positive results: 80 + 270 = 350. C is not correct. 9% chose this. The figure 100 is the number of people in the town with disease X (ie, the prevalence of disease X). This is calculated: 1000 × 0.10 = 100. D is not correct. 12% chose this. The figure 270 is the number of people who will have an incorrect positive screening test result (ie, false positives). One way of calculating this is that there are 900 people without the disease (with a 10% incidence in 1000 people, 100 will have the disease and 900 will not). If the specificity is 70% (the percentage of true-negative test results in people without the disease), then there will be 630 people who are correctly negative (true negatives). This means that there are 900 – 630 = 270 people without the disease that will test positive (false positives). E is not correct. 17% chose this. The figure 630 is the number of people who will have a correct negative screening test result (ie, true negatives). If 100 people have the disease (10% prevalence), then there are 900 who don't have the disease: 1000 – 100 = 900. The test is 70% specific (ie, this is the percentage of true-negative results detected by the test in a population without the disease): 900 × 0.70 = 630.
Bottom Line: Sensitivity is a measure of how well a test is able to identify those with disease. It is calculated by using the formula: sensitivity = TP / (TP + FN).
|
|
|
Post by Admin on Jan 17, 2016 20:08:54 GMT
|
|