Pre & Post test probabilities 

The idea is to see how varies the probability of detecting the illness in the patient after making him a clinical test in its diagnosis. The main data should be had the previous probability to the same one, called: Pre-test. This value can be obtained in several ways: a) an estimate based on the clinical experience, b) the prevalence of the illness, etc. To make this the Theorem of Bayes is used as follows:

Positive Post test Probability:  VPP = Positive Predictive Value

Negative Post test Probability = (1 – Negative Predictive Value) = 1 – VPN

The traditional calculations were based on the following steps: 1) to transform the probabi lities pre test to Odds, 2) to obtain the Post test Odds by multiplying these by the Positive Likelihood ratio; and 3) retransform these Odds to probabilities post test. A shorter way is to use the Fagan’s Nomogram, where one enters with the value of the probability pre test, the Likelihood respective ratio crosses and the probabilities post test are obtained.

By using the present clinical algorithm both procedures can be improved. Only is needed the values of sensitivity and specificity of the clinical method, then they are introduced and two curves that show the variability of the probabilities post test are obtained (in function of the probabilities pre test). Therefore, the clinician can introduce his estimated value of the pre test probability for obtaining the post test probability, just by looking the correspon ding charts. Also given the respective numerical tables are presented.

See Example 1, in page 6.13 (for better explanation to read the whole topic 6.5)

Variability of the clinical indexes with the prevalence

Just as it was explained in the Topic 4, one can see like the variability of the extrinsic clinical indexes with the prevalence of the illness - that is a way to consider to the previous probabilities to the test, or pre test probability.

By using the same algorithm, the intrinsic clinical indexes can be obtained at the beginning and to see the curves of the extrinsic ones at the bottom of the page developed in Excel.  

The application examples can take of those developed in the text.