Interobserver and intraobserver repeatability of lipid layer pattern evaluation by two experienced observers.

Area of Science:

• Ophthalmology
• Optometry
• Tear Film Dynamics

Background:

• The ocular surface's lipid layer is crucial for limiting tear film evaporation.
• Current lipid layer pattern (LLP) evaluation methods rely on subjective interpretation of interference patterns.
• Objective and repeatable assessment of LLP is needed for accurate diagnosis and management of dry eye disease.

Purpose of the Study:

• To compare the consistency of lipid layer pattern (LLP) evaluations between two experienced observers.
• To assess the intraobserver repeatability of LLP categorization over a one-month interval.
• To identify common misinterpretations in LLP assessment.

Main Methods:

• Lipid layer patterns (LLP) were captured using Tearscope-plus and a digital video camera.
• 124 LLP images were categorized by two expert observers using Guillon's scheme in two separate sessions.
• Interobserver and intraobserver repeatability were analyzed using Cohen's kappa coefficient.

Main Results:

• Substantial interobserver agreement was observed, with Cohen's kappa coefficients of 0.615 and 0.633.
• Substantial intraobserver agreement was also found, with kappa coefficients ranging from 0.770 to 0.812.
• The most frequent misinterpretations involved distinguishing between open/closed meshwork and wave/closed meshwork patterns.

Conclusions:

• While experienced observers demonstrate substantial correlation in LLP categorization, occasional misinterpretations occur, even within the same observer.
• Refining the classification scheme by incorporating intermediate patterns could potentially reduce misinterpretations between adjacent categories.
• Standardized training and potentially objective analysis tools may further enhance the reliability of LLP assessment.