Diagnostic Test Terminology And Performance Measures
What Is An In Vitro Diagnostic (IVD)?
In Vitro Diagnostic (IVD) tests are medical devices that are used in vitro for the examination of specimens (including blood, urine and tissue samples) taken from the human body to provide information on diseases, conditions, or infections, and are sold as diagnostic kits to laboratories, health professionals, or consumers for home use.
What Is A Laboratory Developed Test (LDT)?
Laboratory-Developed Tests (LDT) are “in-house” tests that are designed, manufactured and used in a single laboratory, and sold as a service to clinics, patients, and third-party payers.
What Is An ELISA Test?
Enzyme-Linked Immunosorbent Assay (ELISA) is a biochemical method used to detect the presence of an antibody or antigen in a liquid sample. ELISAs are the most widely used assay type, as it is a relatively fast and inexpensive test method.
What Are Tumour Markers?
Tumour markers are substances that are produced by cancer or by other cells of the body in response to cancer, such as proteins, gene expression patterns and DNA fragments. Tumour markers are often found at higher than normal levels in the blood, urine, stool, tumour tissue, or other tissues or body fluids of some people with cancer. However, tumour markers may be elevated, normal or absent in people with cancer, can vary over time, and may be present at an early or later stage of cancer. Tumour markers can be used to aid the detection, diagnosis and management of cancer to help improve patient outcomes and survival.
What Is A ROC Curve?
A receiver operating characteristic (ROC) curve graphically shows the performance of a diagnostic test by plotting the true positive rate (TPR or sensitivity) for correctly detecting cancer in diseased patients against the false positive rate (FPR, or 1-specificity) for incorrectly identifying cancer in non-diseased (healthy) patients, at various threshold settings. A ROC curve can be used to compare diagnostic tests. A good diagnostic test must demonstrate a high sensitivity (correct diagnosis) and acceptable false positive rate (“cancer scares”) for the disease.
What Is The AUC?
The area under the curve (AUC) interprets the accuracy of the test in distinguishing between patients with and without cancer, where the greater the AUC the better the test. A perfect test would have an AUC=1.0, an excellent test AUC=0.9-0.99, a good test AUC=0.8-0.89, and a useless test AUC=0.5. In a clinical context, an AUC=0.90 means the probably of correctly classifying a patient as being positive or negative for cancer is 90%.
What Is Sensitivity?
Sensitivity (or true positive rate) is the percentage of patients with cancer that were correctly identified with a positive test result. High sensitivity is important because it minimises the number of patients with cancer that are missed (false negatives). For example, if a test has a sensitivity of 90% for lung cancer, then it detects 90% of people with lung cancer and misses lung cancer in 10% of people with cancer.
What Is Specificity?
Specificity (or true negative rate) is the percentage of patients without cancer that were correctly identified with a negative test result. For example, if a test has a specificity of 90% for lung cancer, then 90% of healthy people will correctly test negative and the other 10% will be false positives.
What Is A False Positive?
False positive (1 – specificity) is the percentage of heathy people without cancer that were incorrectly identified with a positive test result. A diagnostic test with low specificity for cancer may have an unacceptably high false positive rate (“cancer scares”) that can lead to patient anxiety and over referral of healthy individuals for unnecessary invasive and costly follow-up procedures.
What Is Accuracy?
Accuracy is the percentage of true results. It is an overall measure of test reliability in correctly identifying people with and without cancer. It is calculated by dividing the number of true positive and true negative results by the test population.
What Are Predictive Values?
Positive predictive value (PPV) is the probability that the disease is present when the test is positive in a given population with known disease prevalence.
Negative predictive value (NPV) is the probability that the disease is not present when the test is negative in a given population with known disease prevalence.
For example, in a high-risk population with 20 cases of disease for 1000 people (2% prevalence), a diagnostic test with a specificity of 90% and sensitivity of 90% would have a PPV=15.5% and NPV=99.7%.