Everything all right in method comparison studies?

Abstract

Researchers and clinicians often need to know whether a new method of measurement is equivalent to an established one that is already in use. For this problem, the estimation of limits of agreement advocated by Bland and Altman is a widely used solution. However, this approach ignores two vital issues in method comparisons. First, does the appropriate re-scaling of the test method bring the methods into agreement? Second, independent of lying ‘adequately’ between the limits of agreement or not, it is important to know whether one method is equal to or better than another. This article proposes an approach and a model, where both these questions will be addressed simultaneously. In this model, the error variation of the standard method stands for ‘acceptable’ precision in measurements. Accordingly, the between-subject component of the measurements by the standard method will be used as a ‘gold standard’ against which the properties of the test method will be evaluated. Application of the model is demonstrated using the peak expiratory flow rate data of Bland and Altman.