ABSTRACT
Determining whether a patient is likely to have a specific disease or condition usually begins with an a priori probability for an occurrence. This is often the prevalence (or pretest probability) of the disease in a specific population. Most diagnostic tests are not perfect, but the results of the test(s) will be used to increase or decrease our estimate of the likelihood (posttest probability) of the disease. This process is sometime referred to as the refining probability. The most important reason pharmacists and other health professionals order a test is to help refine probability and make a decision about the best approach to treating the patient. This refining probability is the process of modifying our estimate of the probability that a disease or condition is present through the results observed on some diagnostic test(s). As will be developed in this chapter, probabilities are critical in predicting the likelihood for a particular disease in a given patient. This prediction will be based on the prevalence of the disease and the likelihood ratio associated or a modification of conditional probability resulting from a diagnostic test, which is affected by the test’s sensitivity and specificity. Sensitivity and Specificity Conditional probability was important when we discussed the chi square test of independence. Based on Eq. 2.6 the probability of some level of variable A given a certain level of variable B was defined as
and if the two discrete variables are independent of each other, then the probability of each level of A should be the same regardless of which B characteristic it contains.
