Requirements for analytical performance in clinical chemistry : an evaluation from the point of view of the practicing physician

In today's medicine many decisions are based on data from the clinical chemistry laboratory. Basically laboratory results for a specific constituent serve a dual purpose (35, 65): a. to recognize the diseased individual by relating the value found to either the range of values observed in a healthy population or to a range of values known to be compatible with a specific condition. b. to monitor changes in an individual patient. Consequently measurements in clinical chemistry have to be reliable in two ways. Firstly, they have to be accurate, which is defined (I ,34a) as close to the true value or an accepted reference value and secondly, they have to be precise, which means, that replicate test results have to agree closely (1,34a). The question that has often been discussed but hitherto not been answered satisfactory is: how accurate and how precise do measurements in clinical chemistry have to be? In general in technical disciplines, requirements for accuracy and precision of measurements are set by the consumer needs. For example in steel construction the consequences of measurement errors are carefully considered. If the length of beams is incorrect or shows variations, all or some will not fit. If the actual shapes are smaller than designed shapes or if the strength of the material is insufficient the construction may collapse. If the strength/weight ratio is low the beams will be too expensive to meet the budget. Technical and economical consequences of each deviation from the ideal or required dimensions can be calculated, and, not unimportant, instruments used for measurements are adequate to register these deviations. In clinical chemistry consumer needs are more difficult to define. Firstly, ideal or required values are difficult to establish.Secondly, clinical chemistry measurements are seldom carried out with impeccable accuracy and precision.