Agreement between Automated and Traditional Measures of Tear Film Breakup

Abstract

To determine the repeatability and agreement between the noninvasive Keratograph tear break-up time (NIK-BUT) as measured by the Oculus Keratograph 4 and fluorescein tear break-up time (FBUT). Sixty subjects were recruited for two study visits separated by 7 (± 2) days. At each visit, three NIK-BUT measures and FBUT measures were obtained. Each NIK-BUT measure from the Keratograph included a first and an average NIK-BUT. The means of the measures obtained, first NIK-BUT, and average NIK-BUT and FBUT were calculated for each visit. Between- and within-visit agreement was assessed using intraclass correlation coefficients (ICCs) and Bland-Altman 95% limits of agreement (LoA) analyses of log-transformed data. Between-visit ICCs were 0.53 [95% confidence interval (CI), 0.32 to 0.69] for first NIK-BUT, 0.59 (95% CI, 0.40 to 0.73) for average NIK-BUT, and 0.66 (95% CI, 0.49 to 0.78) for FBUT, whereas 95% LoA were -0.65 to 0.67, -0.44 to 0.48, and -1.14 to 1.10 [back transformed: (visit 1 + 0.01)/(visit 2 + 0.01) = 0.22 to 4.68, 0.36 to 3.02, and 0.07 to 12.59] for the aforementioned methods, respectively. The visit 1 within-visit ICC between first NIK-BUT and FBUT was 0.44 (95% CI, 0.21 to 0.62), whereas the 95% LoA was -0.84 to 1.18 [back transformed: (first NIK-BUT + 0.01)/(FBUT + 0.01) = 0.14 to 15.14]. Likewise, the visit 1 within-visit ICC between average NIK-BUT and FBUT was 0.41 (95% CI, 0.18 to 0.60), whereas the 95% LoA was -0.58 to 1.44 [back transformed: (average NIK-BUT + 0.01)/(FBUT + 0.01) = 0.26 to 27.54]. The 95% LoA suggest that the average NIK-BUT has better between-visit agreement compared with the first NIK-BUT or FBUT. The first NIK-BUT showed better within-visit agreement with the FBUT than the average NIK-BUT. In addition, there is better between- and within-visit agreement for all measures at lower values.