Abstract
An overview is given of the main difficulties encountered when trying to apply the concept of medical usefulness to “analytical goal-setting” in clinical chemistry. As part of the essay the elements of medical decision theory are explained, and previous attempts at answering the questions we are facing are contrasted with the answers or approaches suggested by decision theory. Among the difficulties are the multitude of purposes for which clinical data are collected, the vast number of unknowns and ill-defined elements in medical practice, the limitations of the clinician's mind as processor of information, and the need to take the entire health budget into account when planning research and development in a particular speciality. Clinical chemists should resist the temptation to stick to over-simplified pictures of the use that the clinician makes of diagnostic tests, of the consequences of his decisions for the patient's welfare, and of the desirability of accuracy and precision per se. For several reasons the inventory of tests should not proliferate too fast. Evaluation in terms of tail areas (false positives and negatives, etc.) is faulted for being unrealistic.