Effect of exclusion criteria on the distribution of blood test values

TL;DR

This study investigates how exclusion criteria affect the estimated distributions of blood test values, highlighting their significant impact on the shape of these distributions in health checkup data.

Contribution

It provides a quantitative analysis of how exclusion criteria influence blood test value distributions, aiding in more accurate personalized health assessments.

Findings

Exclusion criteria significantly alter distribution shapes.

Differences in estimated distributions depend on exclusion criteria.

Understanding these effects improves personalized health evaluations.

Abstract

The increasing demand for personalized health care has led to the expectation that individualized quantitative evaluation of human disease states is possible. However, this has not yet been achieved at a sufficiently low cost. Our ultimate goal is to determine the most accurate distributions of blood tests commonly used in health checkups. In this study, we quantified differences between the estimated distributions based on four datasets using the lognormal distribution with three parameters and analyzed the cause of the differences. We focused on two causes of differences: the exclusion criteria and distribution used for estimation of distributions. We compared the expected values across datasets for each laboratory test. We also quantitatively evaluated differences in the shape of the estimated distribution corresponding to the exclusion criteria. We found that exclusion criteria have an important influence on the shape of the distribution for blood test values.