Reliability and clinical applicability of a novel tear film imaging tool

Abstract

Purpose: The aim of our research was to investigate the reliability and clinical applicability of a modern tear film imaging tool by comparing the inter- and intragrader difference. The further goal was to compare the non-invasive tear break-up time (NIBUT) measured with the LacryDiag® device with traditional tear film break-up time (TBUT).

Methods: Comprehensive ophthalmological examination was performed, including LacryDiag® (Quantel Medical, France) (lower tear meniscus height measuring (LTMH), superior and inferior eyelid meibography (MeibS MeibI), interferometry (INT), NIBUT), slit lamp examination, and TBUT. Two independent, well-trained graders selected and analyzed the LTMH, MeibI, MeibS, and INT. The second grader reanalyzed the data 1 month later. Intra- and inter-examiner reliabilities were evaluated using intraclass correlation coefficients (ICC), while for categorical variable, Cohen's kappa statistics were provided. The Bland-Altman plot was used for visualization of the agreement between measurements.

Results: Fifty healthy volunteers were examined. For LTMH both the inter- and intragrader variabilities were excellent. Between two graders, the ICC of MeibI was poor; however, between two graders, the ICC of MeibS was good, and the intragrader variability in MeibI and MeibS was excellent. For the INT, both intra- and intergrading were in fair and moderate agreement, although the intragrader agreement was higher. Comparing the NIBUT and TBUT, the agreement was slight.

Conclusion: Based on our results, examination of a patient during follow-up should be performed by the same examiner, because of the slight agreement. The LacryDiag® is a non-invasive, easy-to-use device, which can examine the tear film and save the recordings for easier follow-up.

Keywords: Dry eye disease; Interferometry; Lower tear meniscus height; Meibography; Tear film break-up time.

Fig. 1

Interferometry, lower tear meniscus height (a), non-invasive tear break-up time (b), and upper eyelid photo for meibography (c) with LacryDiag® device (own source)

Fig. 2

Bland-Altman plots showing the agreement between inter- (a) and intragraders (b) measuring low tear meniscus height. The dots represent the differences from the mean value, the continuous line illustrates the mean value, the dotted line depicts the −1.96 standard deviation (SD), and the broken line is for +1.96 SD. Intercooled Stata for Windows (version 13.0) and Microsoft Excel (version 16.0) were used to create the figure

Fig. 3

Bland-Altman plots showing the agreement between inter- (a) and intragraders (b) measuring Meibomian gland loss in the lower eyelid. The dots represent the differences from the mean value, the continuous line illustrates the mean value, the dotted line depicts the −1.96 standard deviation (SD), and the broken line is for +1.96 SD. Intercooled Stata for Windows (version 13.0) and Microsoft Excel (version 16.0) were used to create the figure

Fig. 4

Bland-Altman plots showing the agreement between inter- (a) and intragraders (b) measuring Meibomian gland loss in the upper eyelid. The dots represent the differences from the mean value, the continuous line illustrates the mean value, the dotted line depicts the −1.96 standard deviation (SD), and the broken line is for +1.96 SD. Intercooled Stata for Windows (version 13.0) and Microsoft Excel (version 16.0) were used to create the figure

Fig. 5

Bland-Altman plots showing the agreement between inter- (a) and intragraders (b) examining interferometry. The dots represent the differences from the mean value, the continuous line illustrates the mean value, the dotted line depicts the −1.96 standard deviation (SD), and the broken line is for +1.96 SD. Intercooled Stata for Windows (version 13.0) and Microsoft Excel (version 16.0) were used to create the figure