AIM: To investigate repeatability and accuracy of a latest Keratograph for evaluating the tear film stability and to compare its measurements with that of traditional examination methods.

METHODS: The results of noninvasive tear film break-up time(NI-BUT)including the first tear film break-up time(BUT-f)and the average tear film break-up time(BUT-ave)were measured by Keratograph. The repeatability of the measurements was evaluated by coefficient of variation(CV)and intraclass correlation coefficient(ICC). Wilcoxon Signed-Rank test was used to compare NI-BUT with fluorescein tear film break-up time(FBUT)to confirm the correlation between NI-BUT and FBUT, Schirmer I test values. Bland-Altman analysis was used to evaluate consistency.

RESULTS: The study recruited 48 subjects(48 eyes)(mean age 38.7±15.2 years). The CV and ICC of BUT-f were respectively 12.6% and 0.95, those of BUT-ave were 9.8% and 0.96. The value of BUT-f was lower than that of FBUT. The difference had statistical significance(6.16±2.46s vs 7.46±1.92s, P<0.01). There was significant positive correlation between NI-BUT and FBUT, Schirmer I test values(P<0.01). The scope of 95% limits of agreement(LoA)was 4.46s in BUT-f and FBUT, while the scope of LoA was 3.64s in BUT-ave and FBUT.

CONCLUSION: Keratograph can provide NI-BUT data that has a better repeatability and reliability, which has great application prospects in diagnosis and treatment of dry eye and refractive corneal surgery.