[Measurement of reliability and validity of clinical instruments]

In biomedical measurement, comparisons between two instruments or between a new instrument and the gold standard is commonly needed to evaluate the reliability and validity of the instruments. Such studies are frequently analyzed using statistics that are inappropriate for studying concordance. This article reviews the appropriate statistical techniques and provides a current methodological guide for the evaluation of reliability and validity of instruments used in the measurement of biomedical variables. The techniques are illustrated using real data from a pilot study comparing four instruments for measuring cholesterol.