How to Establish a Reference Interval: A Practical Approach

Establishing accurate reference intervals (RIs) is essential for interpreting clinical chemistry laboratory test results and ensuring clinical relevance. The aim of this review is to provide a practical guide to determine RIs using direct and indirect methods.

RIs are established by statistical analysis of results from a reference population that is assumed to reflect the distribution of the total population. Two main methods of establishing RIs are the direct method (by active sampling of a population) or the indirect method (by using real-world data). Correct use of a priori and a posteriori exclusion criteria is critical to avoid selection bias. Using the direct method, the choice of the statistical approach to calculate the RI depends on sample size and data distribution. For datasets ≥120 samples, parametric methods are preferred for normally distributed data, whereas nonparametric methods are recommended for skewed distributions.In the case of an indirect method, routine laboratory data can be used, assuming that the pathological fraction is limited or can be identified. Newer indirect methods such as refineR are favored over traditional indirect techniques such as Bhattacharya or Hoffmann. These newer indirect methods are less influenced by atypical reference distributions or larger proportions of pathological data in the analyzed dataset.

This paper outlines a practical approach for establishing RIs, emphasizing the selection of an appropriate reference cohort and the application of the optimal statistical techniques for both direct and indirect methods.