Reference values based on hospital admission laboratory data
S. J. Harwood and G. W. Cole
Laboratory values taken from more than 13,000 newly admitted patients to
the University of Alabama Hospital were analyzed using a percentile ranking
system, and histograms for 24 laboratory tests were prepared. The
percentile system is a better way to express test results than the classic
normal range. The ability to correlate a test result with a clinical
problem rises proportional to the degree of deviation from the midrange.
Minor deviations frequently are without explanation, even after intensive
clinical investigation. Invariably, great deviations are explainable. If
one limits the interpretation of results to three ranges--abnormally low,
abnormally high, and normal--much of the value of laboratory test
interpretation is lost.
