Agreement between two tympanometers. A methodological study of instrument comparison

Area of Science:

• Audiology
• Biomedical Engineering
• Medical Instrumentation

Background:

• Comparing measurements from new and old instruments is crucial for data consistency.
• Correlation analysis is often inappropriately used for assessing instrument agreement.
• A robust statistical method is needed to evaluate measurement agreement and precision.

Purpose of the Study:

• To introduce and validate a simple statistical method for comparing measurements between old and new instruments.
• To assess the agreement and precision of a new digital tympanometer against an older automatic model.
• To evaluate the clinical and scientific significance of measurement discrepancies.

Main Methods:

• A statistical method plotting measurement differences against their means was employed.
• A one-sample t-test was used to assess the distribution of differences against zero.
• Measurements of static compliance, middle ear pressure, and gradient were compared between two tympanometers in 32 adults.

Main Results:

• Static compliance and gradient showed good agreement between the old and new tympanometers (P > 0.05).
• Middle ear pressure exhibited a significant bias due to phase delay, though clinically insignificant.
• The new digital tympanometer demonstrated higher precision compared to the older model.

Conclusions:

• The proposed statistical method effectively assesses instrument agreement and precision.
• While minor biases exist, the new tympanometer offers improved precision for audiological measurements.
• Careful statistical evaluation is necessary when comparing data from different generations of medical devices.