Clinical agreement in quantitative measurements

Area of Science:

• Medical Technology Assessment
• Biostatistics
• Clinical Diagnostics

Background:

• Advancing medical technology necessitates reliable assessment of new diagnostic and therapeutic methods.
• Traditional assessment methods like correlation or mean equality are insufficient for evaluating measurement agreement.
• Two statistical approaches, limits of agreement and intraclass correlation coefficient, are commonly used but have limitations.

Purpose of the Study:

• To describe and compare limits of agreement and intraclass correlation coefficient for assessing measurement agreement.
• To discuss the clinical and statistical advantages and disadvantages of each method.
• To guide investigators in selecting and properly applying appropriate agreement assessment procedures.

Main Methods:

• Review and description of statistical principles behind limits of agreement.
• Review and description of statistical principles behind intraclass correlation coefficient.
• Application and comparison of both methods using a real-world clinical example.

Main Results:

• Correlation and mean equality are inadequate for assessing agreement between measurement methods.
• Limits of agreement provide a range for individual differences, useful for clinical interpretation.
• Intraclass correlation coefficient quantifies reliability and consistency but offers less direct clinical insight.

Conclusions:

• Both limits of agreement and intraclass correlation coefficient have distinct merits and demerits for assessing measurement agreement.
• The choice of method depends on the specific clinical question and desired interpretation.
• Proper application and understanding of these statistical tools are essential for accurate medical technology assessment.