Although the concepts of sensitivity and specificity are usually applied to diagnostic laboratory or screening tests, they are equally valid for other tests that indicate the presence or absence of disease. These may be tests designed to elicit certain physical signs of disease by examination (e.g. hypertension, cervical cancer, neurological defects, and tendency to hemorrhage) or questions aimed at uncovering symptoms.

Suppose that a short psychometric questionnaire that was devised to detect undiagnosed cases of personality disorder was found, after being tested in different mental health clinics using the questionnaire on people with previously diagnosed personality disorder, to have a sensitivity of 25% and a specificity of 99%.

Imagine you are a public health official in charge of deciding whether or not your clinics will use this test to diagnose personality disorder. Discuss why you would (or would not) recommend the use of a test that gave these results (specific to both this actual test, and the type of population, i.e., prevalence of disease)? Hint: Be sure to discuss results relating to sensitivity, specificity, and PPV