Everybody has been exposed to diagnostic testing in his or her lifetime. Diagnostic Testing is typically used to differentiate between people with/without a particular disease.So your doctor orders a panel of tests to determine whether or not you are part of the prevalence population for disease A. Each diagnostic test has a particular accuracy rate of detecting true positive and true negative cases. Another words, you can test positive for a diagnostic test and not have the disease. You can also test negative for something and you have the disease.
These diagnostic tests are suppose to be analyzed to determine the accuracy of the test. The ideal situation is that a positive test result only shows up when you have the disease. You also want a negative test result when you don’t have the disease. That is not always the case. So doctors use statistics to determine specificity and sensitivity of the test. Sensitivity is the probability that a test will detect true positive results. Specificity is the probability that a test will detect true negative results. You want high probabilities for both statistics. A high probability means, there is a high rate of accuracy in the test. It would be great if you obtain a sensitivity approximately equal to your specificity. That is not always the case.
There are other ancillary statistics related to the topic. Negative predictive probability looks at the likelihood of a person not having the disease given their test result is negative. Positive predictive probability is the likelihood of a person having disease given their test result is positive. Now, its a little confusing to retain all of these test. I suggest you think about the focus- that might help. So you are focused on the test performance with sensitivity and specificity. But negative and positive predictive probability focuses on the disease state of the individual.
As a side note, the equation is not the trap here. Your mind has to accept the implications of the probabilities. So even if you memorize it, I encourage you to find examples and work them out.
If you enjoyed this blog, there are more to follow-Amy