Intrasession repeatability and interobserver reproducibility of ICP Tearscope measurements of tear meniscus height and non-invasive breakup time

Abstract

Purpose: To examine intrasession repeatability and interobserver reproducibility of tear meniscus height (TMH) and non-invasive tear breakup time (NIBUT) measurements provided by ICP Tearscope, and their agreement with Keratograph 5 M measurements in persons with and without dry eye.

Methods: Participants were 48 individuals with dry eye disease (DED) and 44 healthy controls. To determine intrasession repeatability, two consecutive TMH and NIBUT measurements were carried out by a single experienced examiner using ICP Tearscope. For interobserver reproducibility, a second masked investigator measured TMH and NIBUT on the first image and video captured by the examiner. Keratograph 5 M measurements of both variables were carried out by the same examiner. Repeatability and reproducibility were evaluated through within-subject standard deviation (Sw), coefficient of variation (CoV), intraclass correlation coefficient (ICC) and Bland-Altman plots (bias and limits of agreement (LoA)).

Results: The TMH measurement with ICP Tearscope showed both, a moderate intrasession repeatability (Sw = 0.045, CoV < 19.82 %, ICC > 0.684) and interobserver reproducibility (Sw = 0.032, CoV < 14.09 %, ICC > 0.926). The NIBUT measurement showed worse intrasession repeatability in the DED group (CoV = 27.53 %) than in the Control group (CoV = 13.31 %). Even though similar Sw, LoA and ICC values were observed in both groups (Sw = 2.55 s and 2.34 s; LoA = -6.60 to 5.18 s and -4.82 to 6.51 s, Control and DED respectively, ICCs > 0.95). Interobserver reproducibility indicated a moderate amount of variability with CoV ≤ 15.01 % recorded in both groups. Agreement between the two devices was poor. Limits of agreement were not clinically acceptable for TMH (LoA = -0.19 to 0.14 mm and -0.16 to 0.13 mm, Control and DED, respectively) and for NIBUT (LoA = -12.74 to 14.08 s and -12.19 to 12.87 s, Control and DED, respectively).

Conclusions: ICP Tearscope provides TMH and NIBUT measurements that show weak repeatability and moderate interobserver reproducibility. This device is not interchangeable with the Keratograph 5 M.

Keywords: Agreement; Non-invasive tear breakup time; Repeatability; Reproducibility; Tear meniscus height; Tearscope.