Contents lists available at ScienceDirect Contact Lens and Anterior Eye journal homepage: www.elsevier.com/locate/clae Evaluating the diagnostic ability of two automated non-invasive tear film stability measurement techniques Joevy Lim, Michael T.M. Wang, Jennifer P. Craig* Department of Ophthalmology, New Zealand National Eye Centre, The University of Auckland, New Zealand ARTICLE INFO Keywords: Dry eye diagnosis ocular surface tear film stability breakup time ABSTRACT Purpose: To evaluate the comparability, discriminative ability, and optimal thresholds for non-invasive tear film breakup time measurements obtained from the Oculus Keratograph 5M and Medmont E300 in detecting other signs and symptoms of dry eye disease, as defined by the TFOS DEWS II diagnostic criteria. Methods: One hundred and thirty-four participants (53 male, 81 female), with a mean ± SD age of 48 ± 20 years, were recruited into a prospective, investigator-masked, diagnostic accuracy study. Dry eye symptomology, tear film parameters (including non-invasive Keratograph and Medmont breakup time), and ocular surface staining were evaluated in a single clinical session. Results: Significant positive correlation was observed between the two automated instruments (p < 0.001), al- though non-invasive breakup time measurements obtained from the Medmont were significantly longer (p < 0.001), and demonstrated greater intra-subject and inter-subject variability (all p < 0.001). The areas under the ROC curves exceeded 0.65 for both instruments, and the discriminative abilities were comparable (p = 0.53). The Youden optimal diagnostic threshold for non-invasive tear film stability measurements obtained from the Keratograph was ≤8 seconds, and the optimal cut-off for breakup time measurements obtained from the Medmont was ≤14 seconds. Conclusions: Despite significant positive correlation, breakup time measurements obtained from the Keratograph and Medmont were not directly interchangeable. Measurements from the Medmont were significantly longer and demonstrated greater intra-subject and inter-subject variability, although the two automated, non-invasive methods for assessing tear film stability exhibited comparable overall performance in diagnosing dry eye disease. 1. Introduction Dry eye disease is among the most common ophthalmic conditions encountered in clinical practice,[1] affecting between 5% to 50% of the adult population in different regions of the world.[2] The condition is characterised by the loss of ocular surface homeostasis, which induces a self-perpetuating cycle of inflammatory cascades, tear hyperosmolarity, and tear film instability.[3,4] The resulting symptoms of ocular dryness and irritation are recognised to have significant impacts on quality of life, vision, and work productivity.[5–9] Tear film stability assessment, through the measurement of breakup time, is an essential component of the diagnosis and monitoring of dry eye disease, and forms part of the global consensus diagnostic criteria recommended by the Tear Film and Ocular Surface Society Dry Eye Workshop II (TFOS DEWS II).[7] Although fluorescein tear film breakup time evaluation has traditionally been used, the destabilising effects and artificially shortened measurements following aqueous fluorescein instillation have been increasingly recognised.[7,10,11] The TFOS DEWS II diagnostic methodology report consequently re- commended the use of automated, non-invasive breakup time mea- surement techniques, in favour of the traditional fluorescein method, where available.[7] The Keratograph 5M (Oculus Optikgeräte GmbH, Wetzlar, Germany) and the Medmont E300 (Medmont International Pty Ltd, Melbourne, Australia) are both commercially available corneal topo- graphic instruments, which utilise dynamic automated videokerato- scopy to measure non-invasive tear film breakup time, through de- tecting distortion in the contours of reflected Placido disc mires. [11–13] To our knowledge, the agreement and comparative diagnostic ability of the tear film stability measurements obtained from the two instruments are unknown. The purpose of this diagnostic accuracy study was therefore to evaluate the comparability, discriminative https://doi.org/10.1016/j.clae.2020.08.006 Received 31 March 2020; Received in revised form 22 June 2020; Accepted 1 August 2020 ⁎ Corresponding author at: Department of Ophthalmology, New Zealand National Eye Centre, The University of Auckland, New Zealand, Private Bag 92019, Auckland 1142, New Zealand. E-mail address: jp.craig@auckland.ac.nz (J.P. Craig). Contact Lens and Anterior Eye xxx (xxxx) xxx–xxx 1367-0484/ © 2020 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. Please cite this article as: Joevy Lim, Michael T.M. Wang and Jennifer P. Craig, Contact Lens and Anterior Eye, https://doi.org/10.1016/j.clae.2020.08.006