Accuracy and precision in clinical chemistry.

Introd u ction Clinical chemistry, for decades restricted to a limited number of chemical tests, has, in recent years, through an increasing awareness of the value of the laboratory investigation of disease, and the introduc­ tion of automation and sophisticated instru­ mentation, expanded its field of operations to include a bewildering variety of tests. Advances in biochemical knowledge have provoked the introduction of new tests, particularly in the domain of enzymology. Yet, paradoxically, the bulk of the work of the present time routine laboratory consists essentially of the same small number of time-honored tests, employing the same chemical principles, as in the past. It is true that there is now a trend towards measur­ ing enzyme activity and even some sub­ strates by rate reaction methods; but, high capital cost of automation and the slowness and expense of changing its methodology have resulted in the main in the fossilization of methods, current when the automa­ tion was conceived but not necessarily the appropriate ones today. It is probable that these methods will be with us for many years to come; with one or two exceptions (mainly enzymes) current automation is performing the traditional tests by essen­ tially the traditional methods. It is there­ fore of paramount importance for us to consider exactly what it is that is being measured and how well the measurements are being made. In addition, we must ask two supplementary questions—does the test measure what we want it to measure? and—how close to the truth must the an­ swer be for it to be useful in the diagnosis and treatment of disease? To answer these questions it is necessary to discuss the meaning of precision and accuracy.