| page1 | page2 | page3 | page4 | page5 | page6 | page7 | page8
|
that, notwithstanding the test result, the cause may be cardiac in origin and he may be at risk for further problems which will need immediate medical attention?" The answer to these questions is found in all of the major textbooks on cardiology, and summarized in various versions of the illustration on page 1, using Bayesian analysis. In the graph, the horizontal axis is the pretest probability of coronary artery disease, which, in this case, was roughly 85 percent. The vertical axis is the posttest probability that coronary artery disease caused the symptoms. The column indicating (+ST) refers to a positive |
test suggestive of ischemia; the column indicating (-ST) refers to a negative test result. The chart clearly demonstrates that when the pretest probability is 85 percent and the treadmill test is negative, the posttest likelihood that the symptoms were caused by coronary artery disease is still greater than 50 percent. The physician who understands the limitations of the treadmill stress test and the principles of Bayesian analysis is never justified in reassuring a patient that typical angina is not caused by coronary artery disease. By doing so in this case, the defendant created a false |
sense of security in Mr. Johnson. Plaintiffs argued that "forewarned is forearmed" and that instead of being forewarned, Mr. Johnson left the hospital disarmed by the poor advice given by Dr. Truong. Bayesian principles apply to many types of clinical and laboratory tests done in many aspects of medicine. In every case where a diagnosis is missed, justified on the basis of a test result, it is important to find out what the sensitivity and specificity of the test is, and how Bayesian analysis comes into play to determine the posttest likelihood of disease, notwithstanding a negative test result. |
|
|
|||
| page1 | page2 | page3 | page4 | page5 | page6 | page7 | page8