Application of the Hoffmann, Bhattacharya, nonparametric test, and Q-Q plot methods for establishing reference...

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Biostatistics

Background:

• Reference intervals (RIs) are crucial for interpreting laboratory test results and guiding clinical decisions.
• Accurate RI estimation is essential for reliable biomarker interpretation.
• Various statistical methods exist for RI determination, each with potential advantages and limitations.

Purpose of the Study:

• To evaluate and compare the application value of four different methods (Hoffmann, Bhattacharya, nonparametric test, and Q-Q plot) for estimating reference intervals.
• To assess the reliability of these methods for common laboratory biomarkers including urea, creatinine, and uric acid (UA).
• To verify the differences in RI estimations across methods using biological variability data.

Main Methods:

• A cross-sectional study utilizing patient data from January 2020 to April 2022.
• RI estimation for urea, creatinine, and UA using Hoffmann, Bhattacharya, nonparametric, and Q-Q plot methods.
• Verification of RI differences using reference change value (RCV%) based on biological variability.

Main Results:

• In the clinical group, most method differences for RI upper limits were within the permissible RCV% range.
• The nonparametric test yielded an RI for urea with a lower limit within the permissible RCV% range.
• In the physical examination group, relative RI differences among methods were generally within acceptable RCV% ranges, with specific exceptions for lower limits and UA RI.

Conclusions:

• The Hoffmann and Bhattacharya methods demonstrate potential for providing reliable reference intervals.
• These methods are suitable for indirect estimation of urea, creatinine, and UA from existing laboratory datasets.
• The findings support the use of Hoffmann and Bhattacharya methods in routine laboratory practice for RI determination.