Interpreting Diagnostic Tests Using Probability