Reference interval studies: What is the maximum number of samples recommended?

Background: Little attention has been paid to the maximum number of specimens for reference interval calculation, i.e., the number of specimens beyond which there is no further benefit in reference interval calculation. We present a model for the estimation of the maximum number of specimens for reference interval studies based on setting the 90% confidence interval of the reference limits to be equal to the analyte reporting interval. Methods: Equations describing the bounds on the upper and lower 90% confidence intervals for logarithmically transformed and untransformed data were derived and applied to determine the maximum number of specimens required to calculate a reference interval for 12 common chemistry and hematology analytes. Results: Maximum sample sizes ranged from 126 to 18,171 and depended on the standard deviation of the population, any transformation involved and on the chosen reporting interval. Conclusions: This paper demonstrates the importance of the influence of reporting interval on reference intervals. Using this technique can reduce the cost of determining a reference interval by identifying the maximum number of specimens required.