Assessing agreement between two methods of measuring blood glucose level without Bland-Altman limits

Area of Science:

• Biomedical Engineering
• Clinical Diagnostics
• Statistical Methods

Background:

• The Bland-Altman method is commonly used for assessing agreement between measurement methods.
• Existing methods for assessing agreement in blood glucose measurements have several limitations.
• There is a need for a more robust and flexible method for evaluating measurement agreement.

Purpose of the Study:

• To introduce a novel method for assessing agreement between two blood glucose measurement techniques.
• To address the deficiencies identified in the Bland-Altman method.
• To propose a simplified approach that bypasses the need for Bland-Altman limits.

Main Methods:

• Investigated the shortcomings of the Bland-Altman method for agreement assessment.
• Developed a new, simple, and nonparametric method for evaluating measurement agreement.
• This method directly utilizes clinical tolerance limits without calculating Bland-Altman limits.

Main Results:

• The proposed method uses clinical tolerance limits to determine agreement based on the percentage of differences within these limits.
• This approach is simple, robust, and nonparametric, offering greater flexibility.
• It allows for variable clinical tolerance limits based on measurement values and supports non-symmetric limits.

Conclusions:

• Assessing agreement between blood glucose measurement methods can be significantly enhanced.
• Directly employing clinical tolerance limits offers a superior alternative to calculating Bland-Altman limits.
• The new method provides a more effective and adaptable approach to evaluating measurement agreement.