|
Clinicians use laboratory tests to help them make choices. Test results may help reduce uncertainty, make a diagnosis (diagnostic testing), or identify patients who are likely to have occult disease (screening). However, test results may increase uncertainty if the tests poorly discriminate between patients with and patients without disease, if the test results are at variance with the clinical picture, or if test results are improperly integrated into the clinical context.
Laboratory tests are imperfect and may mistakenly identify some people who do not have a specific disease as having the disease (false-positive result) or may mistakenly identify some affected people as not having the disease (false-negative result). A test's ability to correctly include or exclude disease depends on how likely a person is to have a disease (prior probability, see below) and the sensitivity and specificity of the test (see below).
In addition to the risk of providing misleading information, laboratory tests consume limited resources, may delay treatment, may induce unnecessary treatment or cause necessary treatment to be withheld, and may place the patient at risk for an adverse event from the test itself.
Although a single test can provide information about a variety of diseases, a test is often used to establish the presence or absence of only one disease. Similarly, many tests provide a quantitative result (eg, blood sugar 120 mg/dL) or one of several results (eg, exercise ECG with < 1, 1 to 2, > 2 mm ST-segment depression) that are defined as positive only when they meet or exceed some established criterion or cutoff point. Such cutoff points can be based on statistics or can be adjusted to reflect the burdens of false-positive and false-negative results. When laboratory tests are used to establish that disease is present or absent and the test results can be only positive or negative, a test's sensitivity and specificity can be defined.
Sensitivity is the likelihood of a positive test result in patients with the disease; it measures how well the test detects the disease. The false-negative rate is the complement of sensitivity (ie, the false-negative rate plus the sensitivity = 100%).
Specificity is the likelihood of a negative test result in patients without the disease; it measures how well the test excludes disease. The false-positive rate is the complement of specificity.
Conditional
probability is the probability that a disease (or event) will occur if another event, test result, or condition is present. Sensitivity and specificity are special types of conditional probabilities, in this case whether a positive or negative result will occur if the disease is present or absent, as defined by some gold standard (often based on histologic, microbiologic, or radiographic criteria that define the presence or absence of disease).
Prior
(or pre-test) probability is the likelihood that a patient has a condition or disease before the test result is known. Prior probability is based on clinical judgment: how strongly do the symptoms and signs suggest the disease is present, what in a patient's history and risk factors support the diagnosis, and how common is the disease in a representative population?
Post-test
probability is the likelihood the condition or disease is present after the test results are known. The extent to which the test results change assessment of the probability of disease depends on the test's sensitivity and specificity.
Multiple
Screening Tests
Patients often must consider whether to be screened for occult disease. The premises of screening are that early detection of disease can improve outcomes in patients with occult disease and that the false-positive results that often occur during screening do not create a burden that exceeds the benefit of early detection. With the expanding array of screening tests available, the implications of a panel of such tests must be considered. For example, a test with a specificity of 95% will result in 5% of patients without the target disease having a false-positive result. If 2 different screening tests were done, each concerning a different occult disease, in a patient who actually has neither disease, the chance that both tests will be negative is 95% × 95%, or 90%; there would be a 10% chance of at least one false-positive result. If 3 different screening tests for 3 unrelated diseases were done, the chance that all 3 would be negative would be 95% × 95% × 95%, or 86%; corresponding to a 14% chance of at least one false-positive result. If 12 different tests for 12 different diseases were done, the chance of at least one false-positive result is 46%! This underscores the need for caution when deciding on a panel of screening tests and interpreting the results.
Last full review/revision November 2005
Content last modified November 2005
| 
