Principles Of Test Selection And

Reliability

Indicates how well repeated measurements of the same relatively stable phenomenon will give the same result, also known as precision. Reliability may be measured for one observer or for more than one observer.

Example: If on several occasions one clinician consistently percusses the same span of a patient's liver dullness, intraobserver reliability is good. If, on the other hand, several observers find quite different spans of liver dullness on the same patient, interobserver reliability is poor.

Validity

Indicates how closely a given observation agrees with "the true state of affairs," or the best possible measure of reality.

Example: Blood pressure measurements by mercury-based sphygmomanometers are less valid than intra-arterial pressure tracings.

Sensitivity

Identifies the proportion of people who test positive in a group of people known to have the disease or condition, or the proportion of people who are true positives compared to the total number of people who actually have the disease. When the observation or test is negative in persons who have the disease, the result is termed false negative. Good observations or tests have a sensitivity of over 90%, and help rule out disease because there are few false negatives. Such observations or tests are especially useful for screening.

Example: The sensitivity of Homan's sign in the diagnosis of deep venous thrombosis of the calf is 50%. In other words, compared to a group of patients with deep venous thrombosis confirmed by phlebogram, a much better test, only 50% will have a positive Homan's sign, so this sign, if absent, is not helpful, since 50% of patients may have a DVT.

Specificity

Identifies the proportion of people who test negative in a group of people known to be without a given disease or condition, or the proportion of people who are "true negatives" compared to the total number of people without the disease. When the observation or test is positive in persons without the disease, the result is termed "false positive." Good observations or tests have a specificity of over 90% and help "rule in" disease, because the test is rarely positive when disease is absent, and there are few false positives.

Example: The specificity of serum amylase in patients with possible acute pancreatitis is 70%. In other words, of 100 patients without pancreatitis, 70% will have a normal serum amylase; in 30%, the serum amylase will be falsely elevated.

Predictive Value

Indicates how well a given symptom, sign, or test result—either positive or negative— predicts the presence or absence of disease.

Positive predictive value is the probability of disease in a patient with a positive (abnormal) test, or the proportion of "true positives" out of the total population tested.

Example: In a group of women with palpable breast nodules in a cancer screening program, the proportion with confirmed breast cancer would constitute the positive predictive value of palpable breast nodules for diagnosing breast cancer.

Negative predictive value is the probability of not having the condition or disease when the test is negative, or normal, or the proportion of "true negatives" out of the total population tested.

Example: In a group of women without palpable breast nodules in a cancer screening program, the proportion without confirmed breast cancer constitutes the negative predictive value of absence of breast nodules.

Displaying Clinical Data. To use these principles, it is important to display the data in the 2 x 2 format diagrammed on the following page. Always using this format will ensure the accuracy of your calculations of sensitivity, specificity, and predictive value. Note that the presence or absence of disease implies use of a "gold standard" to establish whether the disease is truly present or absent. This is usually the best test available, such as a coronary angiogram for assessing coronary artery disease or a tissue biopsy for malignancy.

Note that the numbers related to presence or absence of disease, as determined by the "gold standard," are always displayed down the table in the left and right columns (present = a + c; absent = b + d). Numbers related to the observation or test are always displayed across the table in the upper and lower rows (testpositive = a + b; test negative = c + d).

Invisible Viagara

Invisible Viagara

You are about to discover the "little-known" techniques, tricks and "mind tools" that will show you how to easily "program" your body and mind to produce an instant, rock-hard erection. Learn how to enjoy all of the control, confidence and satisfaction that comes from knowing you can always "rise to the challenge" ... and never have to deal with embarrassment, apologies, shyness or performance anxiety in the bedroom, ever again.

Get My Free Ebook


Post a comment