[Association or agreement between clinical measurement methods?]

Area of Science:

• Biostatistics
• Medical Statistics
• Clinical Research Methodology

Background:

• Method comparison studies are crucial in various scientific fields.
• Inappropriate analysis, particularly using correlation, can lead to erroneous conclusions about agreement.
• Distinguishing between association and agreement is vital for valid interpretation.

Purpose of the Study:

• To emphasize the critical difference between statistical association and method agreement.
• To demonstrate the utility of the difference plot method for assessing agreement.
• To provide guidance on the correct interpretation of bias and limits of agreement.

Main Methods:

• Conceptual explanation differentiating association from agreement.
• Demonstration of the difference plot method with a practical example.
• Discussion on interpreting bias and limits of agreement.

Main Results:

• Correlation coefficients may indicate strong association but not necessarily agreement.
• The difference plot method visually represents agreement and identifies systematic bias.
• Proper interpretation of bias and limits of agreement is essential for method validation.

Conclusions:

• Avoid using correlation alone for method comparison; focus on agreement.
• The difference plot is a valuable tool for assessing agreement and bias.
• Accurate method comparison analysis ensures reliable scientific conclusions.