Indirect Method for Estimation of Reference Intervals of Inflammatory Markers

Area of Science:

• Clinical Chemistry
• Biomarker Analysis
• Statistical Modeling

Background:

• Direct reference interval (RI) estimation faces challenges including resource demands, defining non-diseased populations, and ethical sample acquisition.
• The indirect method for RI estimation (RII) presents an alternative approach to overcome these limitations.

Purpose of the Study:

• To estimate reference intervals (RIs) for inflammatory biomarkers using an indirect method (RII).
• To evaluate the influence of measurement size and data characteristics on the accuracy of RII.

Main Methods:

• Retrieved C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and presepsin (PSEP) data from a single hospital (April 2009-April 2021).
• Applied Box-Cox transformation for right-skewed data and latent profile analysis to model mixed non-diseased and diseased populations.
• Estimated RIs at the intersection point of distribution curves and evaluated measurement size influence.

Main Results:

• Estimated RIs using RII: CRP 0-4.1 mg/L, ESR 0-10.2 mm/hr, PSEP 0-411 pg/mL.
• Stable results were observed for measurement sizes ≥2,500.
• An abnormal-to-normal value ratio of 0.5 yielded the most accurate CRP results, with specific adjustment values for different measurement sizes.

Conclusions:

• The proposed indirect method (RII) offers a valuable approach for reference interval verification and estimation.
• The RII method provides additional insights, though some limitations should be considered.