But could such a less selective group, more comparable to a practical clinical setting, be used to establish sensitivity and specificity for challenge testing? Unfortunately, there is no gold standard for the diagnosis of bronchial asthma, as there is for coronary artery disease or pulmonary embolism, where coronary and pulmonary angiography are considered diagnostic. If a heterogeneous, overlapping population base were used to establish sensitivities and specificities for challenge testing, there would be doubt as to the diagnosis in many cases, and the values of these parameters would therefore be suspect.
How can pretest probability be estimated? It is very unlikely that precise information on the prevalence of asthma in the population in question would be available, and even if it were, it would be a very crude guide to pretest probability for that patient. The pretest probability should be based on the history, physical examination, and laboratory data for the individual patient. Inspection of Figure 1 shows that this estimate of pretest probability is crucial at low probability levels for positive tests and at high probability levels for negative tests. For positive tests in patients with a likelihood of asthma of at least 50 percent, the exact estimate is not critical, nor is the exact estimate critical in patients with negative tests and a low pretest likelihood. buy female viagra
Granted that the curves cannot be applied in a precise or literal manner to another data base or different methodology, the principles underlying their construction and use are valuable as a guide to clinical evaluation. Test results are often considered in all or none context, or precise cut off points for or against diagnoses are sought. The concept of pretest and post- test probability does not function by deciding unequivocally whether the patient does or does not have a specific disease (in this case bronchial asthma). Instead, this type of approach uses the test result to raise or lower the probability of the disease, and gives some quantitative but imprecise information about how much the probability has been changed by the test.
The concept of pretest and post-test probability stresses the importance of knowledge about the patient when interpreting test results. This may seem self evident, and the guidelines noted above may seem simple common sense. However, as pulmonary function testing becomes commonplace, pulmonologists find themselves interpreting test results without knowledge of the patient. Computerized interpretation of pulmonary function test results, now also becoming commonplace, is the ultimate in this “depersonalized” evaluation. We hope that the concepts discussed in this report will bring a greater perspective and humility to interpretation of pulmonary function data.