Quality specifications for ordinal scale measurements with multiproperty (multiple) urine test strips

Area of Science:

• Clinical Chemistry
• Medical Diagnostics
• Biostatistics

Background:

• Analytical quality specifications are crucial for reliable medical diagnostics.
• Ordinal scale measurements are commonly used in clinical settings, particularly for urine test strips.
• Existing quality specifications often do not adequately address the nuances of ordinal data.

Purpose of the Study:

• To propose novel analytical quality specifications for ordinal scale measurements.
• To provide criteria for multiproperty urine test strips using ordinal scales.
• To establish methods for evaluating trueness, concordance, and repeatability of ordinal data.

Main Methods:

• Utilizing upper limits of healthy reference intervals for criteria development.
• Applying sensitivity and specificity concepts with a grey zone for trueness evaluation.
• Employing Kappa statistics to assess concordance (agreement) between ordinal categories.
• Using binomial statistics for calculating repeatability (precision) of categorized results.

Main Results:

• Established criteria for multiproperty urine test strips based on ordinal scales.
• Defined methods for evaluating trueness, concordance, and repeatability.
• Demonstrated the applicability of statistical tests for ordinal data analysis.

Conclusions:

• The proposed specifications provide a framework for assessing the analytical quality of ordinal scale measurements.
• These criteria enhance the reliability and interpretability of multiproperty urine test strip results.
• The study addresses a gap in analytical quality specifications for ordinal data in clinical diagnostics.