Validation and comparison of five data mining algorithms using big data from clinical laboratories to establish reference intervals of thyroid hormones for older adults

Aim: To establish Reference intervals (RIs) of thyroid-related hormones in older adults using five data mining algorithms and to assess the applicability of each algorithm. Methods: RIs for thyroid-related hormones in older adults were established using five data mining algorithms based on physical examination and patient data. The results of these algorithms were compared to those of RIs established using healthy older adults recruited following strict inclusion and exclusion criteria. The bias ratio (BR) matrix was used to compare the limits of RIs established using different algorithms. Results: Consistency across different algorithms in physical examination data was found to be greater than that of outpatient data. The transformed Hoffmann, transformed Bhattacahrya, kosmic and refineR algorithms showed good performance in calculating RIs from physical examination data. The RIs of Thyroid Stimulating Hormone (TSH) established using Expectation maximization (EM) and patient data were highly consistent with the RIs established using data from healthy older adults. Conclusion: This study recommends the use of transformed Hoffmann, transformed Bhattacahrya, kosmic, and refineR algorithms which are based on physical examination data to establish RIs for thyroid-related hormones in older adults. However, if patient data is used, then an EM algorithm combined with Box-Cox transformation is recommended for data with obvious skewness.