Issues and pitfalls in method comparison studies.

Area of Science:

• Biostatistics
• Medical device validation
• Measurement science

Background:

• Method comparison studies are crucial for validating new measurement techniques against established reference methods.
• Examples include validating non-invasive blood pressure monitors.
• Proper statistical methodology ensures reliable validation.

Purpose of the Study:

• To provide a comprehensive overview of statistical approaches for method comparison studies.
• To critically evaluate common methods like Bland-Altman plots.
• To propose advanced statistical models for improved analysis.

Main Methods:

• Review of classical Bland-Altman plots and their limitations.
• Discussion of alternative statistical methods (correlation, regression, sensitivity/specificity).
• Introduction of linear mixed effects models as a flexible analytical tool.

Main Results:

• Classical Bland-Altman plots have limitations, especially with non-constant bias or multiple observations.
• Other statistical methods like correlation and regression are often misused in this context.
• Linear mixed effects models offer a more adaptable framework for analyzing method comparison data.

Conclusions:

• Linear mixed effects models provide a superior approach for method comparison studies.
• Careful selection of statistical methods is essential for accurate validation of measurement techniques.
• This work guides researchers in choosing appropriate statistical tools for method validation.