Laboratory Diagnosis and Geriatrics : More Than Just Reference Intervals for Older Adults

Objectivity as a Basic Principle of In Vitro Diagnosis

Since antiquity, the in vitro diagnosis has followed the same principle. A biologic sample such as blood or urine collected from the patient would be analyzed for its physical and/or chemical properties, and the result would then be interpreted in terms of clinical information indicating the health status of the patient.

If the procedure is properly calibrated, the results can be expressed quantitatively and described not in words, as is the case for other diagnostic disciplines, such as pathology or radiology, but as an exact numeric value. Thus, the result of a laboratory analysis is not a descriptive observation derived from a subjective experience of the operator, but is an objective measurement, which always is related to a deductive postulated standard. Because most of the results provided by laboratory investigations can be considered as data that are generally reproducible and comparable, laboratory analysis represents a valid instrument of modern, evidence-based medicine.

To keep the clinical interpretation of these data objective, as much as possible, every numeric test result is usually completed by a reference interval, which is listed on the same line next to the test result value. The reference intervals have been derived from the standard statistical distribution of test results obtained from a demographic sample drawn from the normal adult population; normal is understood to be synonymous with healthy.

However, because the statistical reference interval is based on the 95th percentile of Gaussian normal distribution (the bell curve), 5% of results obtained from healthy individuals will always be outside the reference range yet have no pathologic correlation. It follows that even in the case of a healthy individual, the probability that a random test result value would fall outside the reference interval proportionally increases with the number of the same laboratory tests performed on that same healthy individual.

Thus, ironically, the larger the number of tests undertaken, the higher the probability of a patient not being found normal, which shows that the statistically based reference concept has its limits. Every person is not just an anonymous member of a collective, but a unique unrepeatable individual.