Abstract
Normal ranges are essential for the interpretation of laboratory data. In order to recognise pathological processes early in the course of a disease these ranges need to be efficient. The efficiency of the ranges used in the laboratory of the Royal Sussex County Hospital was questioned when biochemical profiling was introduced routinely using the Vickers M300 Multichannel analyser. The number of abnormal results was large and it was apparent that the normal ranges in use at that time, which were designed for a single test performed on a specially selected patient population, were not adequate for screening purposes. The determination of new ranges however, posed several conceptual and practical problems which required resolution. Since there is no universally acclaimed method for determining these ranges several methods requiring simple mathematics and which appeared potentially useful for a laboratory without sophisticated computerisation, were evaluated. The effects of age and sex on these ranges were also investigated. Data from the hospital patient population were used in this work and the effects of data selection on the normal ranges was also evaluated.