Abstract
To assess the adoptability of a new measuring method when true values are unknown, it is necessary to compare the proposed with the conventional method. Although Bland-Altman correlation analysis is widely accepted, it is occasionally misused or applied where it is inappropriate. We provide an overview of Bland-Altman analysis and teach its application and the interpretation of its results. We also comment briefly on regression analysis, another comparison method.
Keywords: Bland-Altman analysis, Limit of agreement, Method comparison, Pitfall, Statistics.