Intact Parathyroid Hormone Reference Intervals: Importance of Pre-Processing for Indirect Reference Interval...

Area of Science:

• Clinical Chemistry
• Biostatistics
• Laboratory Medicine

Background:

• Reference intervals are crucial for interpreting laboratory test results.
• Indirect methods for deriving reference intervals are increasingly used, especially for complex analytes like intact parathyroid hormone (iPTH).
• Data quality and pre-processing significantly influence the reliability of these derived intervals.

Purpose of the Study:

• To evaluate the impact of various data pre-processing strategies on indirectly derived reference intervals for iPTH.
• To compare the effectiveness of different indirect estimation methods (RefLim, TML, refineR) after data pre-processing.
• To assess the utility of vitamin D status and estimated glomerular filtration rate (eGFR) as partitioning criteria.

Main Methods:

• Routine laboratory data for iPTH from a Norwegian site were utilized.
• Data underwent several pre-processing steps: filtering by number of results per patient, exclusion of hospital wards, and biomarker-based filtering.
• Indirect reference intervals were estimated using RefLim, truncated maximum likelihood (TML), and refineR methods.
• Conventional parametric and nonparametric methods were used for comparison.

Main Results:

• Data pre-processing strategies markedly influenced the estimated iPTH reference intervals.
• Filtering for a single measurement per patient effectively excluded pathological cases.
• Excluding specific hospital wards or using biomarker-based filtering had minimal additional impact.

Conclusions:

• Stringent data pre-processing and large datasets are vital for reliable indirect reference interval estimation of iPTH.
• The optimal pre-processing strategy is analyte-specific and requires careful consideration of the analyte's characteristics.
• Further research may be needed to define best practices for different analytes.