Agreement and repeatability of four different devices to measure noninvasive tear breakup time (NIBUT)

Purpose: Since tear film stability can be affected by fluorescein, the Dry Eye Workshop (DEWSII) recommended non-invasive measurement of tear breakup time (NIBUT). The aim of this study was to investigate the agreement and repeatability of four different instruments in the measurement of NIBUT. Methods: 72 participants (mean 24.2 +/- 3.6 years) were recruited for this multi-centre, cross-sectional study. NIBUT was measured three times from one eye using each of the instruments in randomized order on two separate sessions during a day, separated by at least 2 h. NIBUT was performed at three sites (Switzerland, Germany and UK) using three subjective instruments, Tearscope Plus (Keeler, Windsor, UK) (TS), Polaris (bon Optic, Lubeck, Germany) (POL), EasyTear Viewplus (Easytear, Rovereto, Italy) (ET) and the objective Keratograph 5 M (Oculus Optikgerate GmbH, Wetzlar, Germany) (KER). As the latter instrument only analyses for 24 s, all data was capped at this value. Results: NIBUT measurements (average of both sessions) between the four instruments were not statistically significantly different: TS (median 10.4, range 2.0-24.0 s), POL (10.1, 1.0-24.0 s), ET (10.6, 1.0-24.0 s) and KER (11.1, 2.6-24.0 s) (p = 0.949). The objective KER measures were on average (1.2 s +/- 9.6 s, 95 % confidence interval) greater than the subjective evaluations of NIBUT with the other instruments (mean difference 0.4 s +/- 7.7 s, 95 % confidence interval), resulting in a higher limits of agreement. The slope was -0.08 to 0.11 indicating no bias in the difference between instruments with the magnitude of the NIBUT. Repeated measurements from the two sessions were not significantly different for TS (p = 0.584), POL (p = 0.549), ET (p = 0.701) or KER (p = 0.261). Conclusions: The four instruments evaluated for their measurement of tear stability were reasonably repeatable and give similar average results.