Laboratory reference intervals - history and modern approaches for improved utility.

Abstract

Diagnosing using laboratory results is commonly based on whether the concentration of a biomarker is outside of the reference limits following the currently dominating statistical decision theoretical approach. Such dichotomization disregards several crucial factors, including the uncertainty of the measurement result, how extreme the concentration of the biomarker is, the prevalence of the condition in the tested population, and the costs of false positives and false negatives. The history, properties, pros, and cons of reference intervals are discussed, including promising alternatives such as standard scores, percentiles, and likelihood ratios.