Principles of assessment from a clinical perspective

Physicians apply assessments every day in clinical practice. Common clinical practice is impossible without measurements and tests. Testing looks rather straightforward: a test result is either positive or negative. Unfortunately, this simplicity is not in keeping with truth. At the base of measuring and testing in clinical practice lies the assumption of uncertainty: we do not know whether a patient has a disease, we can only estimate the probability that he has a disease by performing a (chain of) test(s). Every test result leaves open the chance that a wrong decision is taken on the basis of the test result. It is a challenge for the clinician to get a better insight into this process, as well as to minimize the chance of wrong decisions. By using a number of clinical examples, we describe here the principles of assessment from two different perspectives: the perspective of the test, and the perspective of the individual patient. The former perspective incorporates test-specific characteristics, such as sensitivity, specificity, accuracy and cut-off levels, and the latter deals with individual probabilities from a 'Bayesian' concept