On the transportability of linear discriminant functions

Area of Science:

• Clinical Chemistry
• Biochemical Diagnostics
• Medical Laboratory Science

Background:

• Analytical methods for biochemical analytes can produce discrepant results.
• Discrepancies in results pose challenges when applying diagnostic algorithms, such as linear discriminant functions, across different laboratories.
• The differential diagnosis of hypercalcemia using alkaline phosphatase activity is a relevant clinical scenario affected by method variability.

Purpose of the Study:

• To evaluate the suitability of alkaline phosphatase (ALP) activity results obtained by a standardized method for use in linear discriminant functions (LDFs).
• To assess the impact of ALP measurement variability on the differential diagnosis of hypercalcemia.
• To determine the feasibility of transporting LDFs between laboratories considering analytical method differences.

Main Methods:

• Utilized results from the Scandinavian Committee on Enzymes (SCE) recommended method for alkaline phosphatase activity.
• Calculated discriminant functions for the differential diagnosis of hypercalcemia.
• Analyzed factors influencing the discriminatory power of the functions.

Main Results:

• Alkaline phosphatase activity results from the SCE-recommended method are suitable for calculating discriminant functions for hypercalcemia diagnosis.
• The chosen method for ALP analysis does not preclude the successful application of LDFs.
• Identified other critical factors that can influence diagnostic discrimination.

Conclusions:

• Linear discriminant functions can be reliably transported between laboratories, provided careful consideration is given to the analytical methods used.
• Standardized methods, like the SCE-recommended ALP assay, facilitate the inter-laboratory transport of diagnostic algorithms.
• Successful implementation requires awareness of potential confounding factors affecting diagnostic accuracy.