18 Chapter 1 History taking and physical examination can also be considered as tests but are outside the focus of this thesis due to their general nature and routine application. Additionally, tests unrelated to healthcare are also outside the scope of this thesis. Such tests include e.g. weight and muscle measurements in gyms, or genealogy tests to trace one’s ancestors. Scientific evaluation of a test To assess the value of a healthcare related test, different aspects should be taken into account [46]: - Analytical performance - Clinical performance - Clinical effectiveness - Cost-effectiveness - Broader impact These concepts are elaborated on in box 1. Box 1. Components of test evaluation Analytical performance: this refers to the ability of the test to accurately detect or measure a particular measurand. Parameters of analytical performance include: trueness: the determination whether the test measures the variable of interest precision: the assessment of the reproducibility of the test. detection limits: a test might not detect a measurand below or above a certain level or might not be specific enough. cross-reactivity: the influence of factors on the test result beyond the measurand of interest. Clinical performance: this refers to the ability of a test in correctly classifying individuals with and without the target condition (such as a disease). This is also called the diagnostic accuracy of a test. Parameters of diagnostic accuracy can be established by comparing the index and reference tests. The index test is the test of interest, while the reference test (also known as the reference standard) is the test to which the index test is compared. The reference test can be the gold standard, but also other options (such as the test in usual care/practice) are used. Clinical performance measures can be obtained by categorizing people with and without the target condition according to their test results in a 2x2 table (table 2): Table 2. Clinical performance of a test in a 2x2 table People with the target condition People without the target condition Total Positive test result TP FP TP+FP Negative test result FN TN FN+TN Total TP+FN FP+TN Total TP: true positives, FP: false positives, FN: false negatives, TN: true negatives Such a table provides insight into the numbers of true positive (TP), true negative (TN), false positive (FP) and false negative (FN) test results. A test can have an inconclusive result as well. Other frequently used parameters of the clinical performance of a test include: sensitivity: the probability of getting a positive test result in people with the target condition (TP/(TP+FN))
RkJQdWJsaXNoZXIy MTk4NDMw