Researchers are developing an enzyme-linked immunosorbent assay test for diagnosing rheumatoid arthritis. The test is designed to detect the presence of serum antibodies against citrullinated proteins. Two test populations with a differing prevalence of rheumatoid arthritis are selected. The researchers plan to assess the test's performance in the 2 populations by comparing a number of diagnostic test parameters. Which of the following performance measures is most likely to be different between the 2 test populations?
Show Explanatory Sources
Predictive values are performance measures of diagnostic tests that are dependent on the prevalence of disease in a population of interest. The positive predictive value (PPV) is the probability that someone who tests positive actually has the disease. It is calculated by dividing the number of true-positive results by the total number of positive results (ie, TP / [TP + FP]). The number of true-positive results depends on the sensitivity, and the number of false-positive results depends on the specificity; the relative proportion of each is determined by the prevalence of disease in the population.
Populations with a lower disease prevalence have fewer true-positive results and higher numbers of false positives, so the PPV decreases. As disease prevalence increases, the number of true positives also increases, while the number of false positives decreases, resulting in a higher PPV. Similarly, the negative predictive value increases as the disease prevalence decreases.
(Choices A and B) Positive and negative likelihood ratios indicate how a particular positive or negative test result influences the pretest probability of having a disease. Likelihood ratios >1 indicate that the test result is associated with the presence of the disease; likelihood ratios <1 mean that the test result is associated with the absence of the disease. Because positive and negative likelihood ratios are based on sensitivity and specificity, they are not affected by disease prevalence.
(Choices D and E) Sensitivity and specificity are not affected by disease prevalence. This is because sensitivity is calculated using true positives and false negatives (only people with the disease), and specificity is calculated using true negatives and false positives (only people without the disease). In this case, the same test (with the same sensitivity and specificity characteristics) has been used on 2 different populations with a differing prevalence.
Educational objective:
Various parameters are used to evaluate the accuracy and usefulness of diagnostic tests. Positive and negative predictive values are influenced by disease prevalence in the target population; sensitivity, specificity, and likelihood ratios are not prevalence-dependent.