Assessing alignment between functional markers and ordinal outcomes based on broad sense agreement

Area of Science:

• Biostatistics
• Clinical Diagnostics
• Medical Informatics

Background:

• Functional markers are crucial for disease diagnosis, requiring assessment of their quantitative features against disease severity.
• Existing methods may not fully capture the alignment between continuous functional marker data and ordinal disease severity scales.
• Broad Sense Agreement (BSA) offers a novel approach to evaluate this relationship.

Purpose of the Study:

• To introduce and validate a framework for assessing the diagnostic utility of functional markers using Broad Sense Agreement (BSA).
• To evaluate the alignment between quantitative features of functional markers and ordinal gold standard tests reflecting disease severity.
• To develop a method for comparing different summary functionals (SFs) for their importance in predicting ordinal outcomes.

Main Methods:

• Adopted a general class of summary functionals (SFs) to capture various quantitative features of functional markers.
• Applied the Broad Sense Agreement (BSA) concept to study the alignment between SFs and ordinal outcomes.
• Utilized three special classes of SFs: AUC-type, magnitude-specific, and time-specific.
• Developed a consistent and asymptotically normal BSA estimator and an inferential framework for comparing SFs.

Main Results:

• The proposed BSA estimator is statistically consistent and asymptotically normal.
• Simulation studies confirmed satisfactory finite-sample performance of the developed framework.
• The methods were successfully demonstrated using a renal study dataset.

Conclusions:

• The proposed BSA framework provides a robust tool for assessing the diagnostic utility of functional markers.
• The framework enables quantitative comparison of different marker features against disease severity.
• This approach has significant implications for clinical studies and disease diagnosis.