Application of Bayes' theorem to the diagnosis of ankylosing spondylitis from radioisotope bone scans.

The ratio of the uptake of radioactivity in each sacroiliac joint to the uptake in the sacrum has been measured in 57 patients with early ankylosing spondylitis and in 51 control subjects. The distribution of ratios of uptake obtained in each of these two groups shows appreciable overlap, and it is shown that the specification of a 'normal range' in this sort of situation can be misleading in the interpretation of the uptake ratio obtained in a given subject. An alternative approach described here is to use Bayes' theorem, which combines a probability based on the numerical value of the measured ratio with the pretest clinical impression about the likelihood of ankylosing spondylitis to yield a post-test probability of the presence of the disease.