Intra-observer repeatability assessment of the S390L Firefly WDR slit lamp in patients with dry eye disease: Objective, automated and non-invasive measures

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published in Eye & Contact Lens: Science and Clinical Practice on 2023, available at: https://doi.org/10.1097/ICL.0000000000001001.” Title Intra-observer repeatability assessment of the S390L Firefly WDR slit lamp in patients with dry eye disease: Objective, automated and non-invasive measures. Running short title S390L slit lamp: Intra-observer repeatability. Authors Antonio Ballesteros-Sánchez, O.D., M.Sc., Ph.D. Candidate. Beatriz Gargallo-Martínez, O.D., Ph.D. Ramón Gutiérrez-Ortega, M.D., Ph.D. José-María Sánchez-González, O.D., Ph.D. Authors Affiliations From the Department of Physics of Condensed Matter (A.B.S., J.M.S.G), Optics Area, University of Seville, Seville, Spain; Department of Ophthalmology (A.B.S., B.G.M., R.G.O), Clínica Novovisión, Murcia, Spain; and Department of Ophthalmology, Optometry, Otorhinolaryngology and Anatomic Pathology (B.G.M., R.G.O), University of Murcia, Murcia, Spain. Corresponding author Antonio Ballesteros-Sánchez, OD, MSc, PhD Candidate. Reina Mercedes Street. University of Seville, Seville, Spain. Telephone: +34 617 70 05 30. E-mail: an[email protected] Conflicts of Interest and Source of Funding This study was funded by the VI University of Seville's Own Research Plan [English editing services, grant number: 8B83-E8F8-1089-414B-A739]. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication. Number of tables and figures: Tables: 2; Figures: 3. Date of submission: November 19, 2022. ABSTRACT Objectives: To assess the intra-observer repeatability of automated, objective and non-invasive measures obtained with the S390L Firefly WDR slit lamp. Methods: This cross-sectional study included fifty eyes of patients with dry eye disease (DED) with a mean age of 55.06 ± 12.96 years. Three consecutively repeated measures of the following variables were obtained: first non-invasive break-up time (F-NIBUT), average non-invasive break-up time (A-NIBUT), tear meniscus height (TMH), tear meniscus area (TMA), nasal ciliary hyperemia (NCIH), temporal ciliary hyperemia (TCIH), nasal conjunctival hyperemia (NCOH), temporal conjunctival hyperemia (TCOH), upper loss area meibomian gland (U-LAMG), lower loss area meibomian gland (L-LAMG), upper meibomian gland dysfunction grade (U-MGD grade) and lower meibomian gland dysfunction grade (L- MGD grade). Intra-observer repeatability was estimated with coefficient of variation (CoV), intra-subject standard deviation (Sw) and Bland-Altman plots. Results: All variables showed no statistically significant differences in the repeated measures analysis except for L-MGD grade (P = 0.045). F-NIBUT and A-NIBUT obtained the highest CoV with an average value of 0.48 ± 0.41 [0.02 to 1.00] and 0.34 ± 0.25 [0.02 to 1.00], respectively. The remaining variables showed CoVs between 0.04 ± 0.11 [0.00 to 0.43] and 0.18 ± 0.16 [0.00 to 0.75]. A-NIBUT, TMA, NCOH and L-LAMG obtained a Sw of 2.78 s, 0.21 mm2, <0.001 and 4.11 %, respectively. Bland-Altman plots showed a high level of agreement between pairs of repeated measures. Conclusion: The S390L Firefly WDR slit lamp has moderate intra-observer repeatability for F-NIBUT and A-NIBUT, which suggests that they are tests with high variability. The remaining variables show satisfactory intra-observer repeatability. Key words Automatic objective measures—Dry eye diagnostic system—Intra-observer repeatability—Meibomian gland dysfunction. INTRODUCTION The definition of dry eye has evolved from its first definition, published in 1995.1 "Multifactorial disease", "loss of homeostasis", "hyperosmolarity" and “neurosensory abnormalities” are the main new concepts included in the current definition by the Tear Film and Ocular Surface Society Dry Eye Workshop (TFOS DEWS) II.2 The prevalence of dry eye disease (DED) varies according to the definition used and the characteristics of the population studied, but it ranges from 5% to 50%, is more frequent in women and increases with age, although it is becoming more frequent in younger people due to the increase in time spent on screens and the use of contact lenses.3 Due to the multifactorial etiology of DED and the lack of correlation between signs and symptoms in DED patients,4,5 there is no reliable gold standard to definitively diagnose and treat DED.6 In addition, traditional diagnostic tests, such as fluorescein tear break-up time (FBUT), Schirmer test (ST) and ocular surface staining, are subjective and invasive tests that depend on the skill of the examiner and influence tear film stability, which can in turn affect test results.6–8 For this reason, non-invasive tests are preferred in the evaluation of DED patients.9 New devices for the assessment of DED have been developed.10 These devices perform objective and noninvasive tests, which reduce observer bias and do not alter tear film stability, resulting in a potential screening tool for DED.11 In addition, combining these tests for a comprehensive ocular surface evaluation, which may include tear film parameters, ocular hyperemia, and meibomian glands imaging, may increase the sensitivity and specificity to diagnose DED and manage its evolution after treatment.12,13 Therefore, it is important to perform studies that assess intra-observer and inter-observer repeatability to validate and demonstrate the reliability of these instruments. Intra-observer repeatability refers to the difference between repeated measures of the same variable obtained by one observer. In contrast, inter-observer repeatability is defined as the difference between measures of the same variable obtained by different observers. To our knowledge, few studies have analyzed the repeatability of the instruments designed for DED diagnosis, and most of them have only analyzed the non-invasive tear break-up time (NIBUT), tear meniscus height (TMH), lipid layer thickness (LLT) and meibography.14–23 In addition, although these new devices perform objective and non-invasive measures, they include some tests that require manual measurement or subjective interpretation by the examiner, such as TMH14,15,19,21,22,24 and meibography,25 respectively. The S390L Firefly WDR slit lamp (Shanghai Mediworks Precision Instruments Co. Shanghai, China) is a new US Food and Drug Administration (FDA)-approved device for the diagnosis of DED. This device includes a dry eye module designed to perform objective and non-invasive measures, which are analyzed by an artificial intelligence (AI) identification system. In addition, this device is able to start some tests automatically with minimal observer intervention. Consequently, the aim of this study is to assess the intraobserver repeatability of the S390L Firefly WDR slit lamp for all automated, objective and non-invasive measures. METHODS Study design and participants This cross-sectional study was performed at the Novovision Ophthalmology Clinic between June 2022 and August 2022. It fulfilled all of the requirements of the Declaration of Helsinki and was approved by the clinical research ethics committee of the University of Murcia (CREC). Prior to initiating the study, informed consent was obtained from each participant. The inclusion criteria were as follows: (1) age ≥ 18 years old; (2) DED diagnosis according to DEWS II,6 meeting one of the following conditions: (2.1) Ocular Surface Disease Index (OSDI) score ≥ 13;26,27 (2.2) NIBUT < 10 seconds; and (2.3) ocular surface staining with > 5 or 9 corneal or conjunctival stains, respectively, and (3) Meibomian gland dysfunction (MGD) diagnosis according to the international workshop on MGD,28 meeting two of the following conditions: (3.1) irregularity of the eyelid margin or mucocutaneous junction; (3.2) vascularity of the eyelid margin; (3.3) plugged or capped Meibomian gland orifices; (3.4) Meibomian gland atrophy; and (3.5) decreased meibum quality and quantity. The exclusion criteria included: (1) all corneal disorders that prevent diagnostic tests from being performed, such as: (1.1) active corneal infections and (1.2) corneal dystrophies; (2) active ocular allergy; (3) pregnant or lactating women; and (4) patients who did not understand or comprehend the informed consent. Patients were instructed not to wear contact lenses one week before the examination, as well as to discontinue the use of eye drops three days before and during the examination to ensure that all repeated measures were performed under the same conditions. Systemic or ocular diseases, previous systemic or ocular treatments and ocular surgeries with more than 6 months of postoperative evolution were not considered exclusion criteria to better reflect the patient population in which this device is to be utilized. Device and measurement protocol The S390L Firefly WDR slit lamp (Mediview R3.0 software) (Shanghai Mediworks Precision Instruments Co. Shanghai, China) includes a dry eye module that performs a comprehensive ocular surface assessment through the combination of objectives and non-invasive tests, which are analyzed automatically by an AI identification system. These tests allow a quick evaluation of tear film stability and volume, ocular hyperemia and Meibomian glands. Regarding technical data, the S390L Firefly WDR slit lamp includes a high-resolution camera with five megapixels and focus calibration that can be manual or automatic. The device also includes a Placido disc that projects 23 ring patterns onto the surface, allowing an examination scope of 8 mm corneal diameter. The Placido disc is located in the slit lamp tonometer hall. Furthermore, a built-in infrared illumination system provides a larger scope capture of Meibomian glands due to an adjustable depth of focus and aperture. Representative images from each test are shown in Figure 1. Tear film stability and volume Tear film stability and volume were evaluated with a Placido disc. When the ring patterns are focused on the ocular surface and the patient blinks twice, the device starts recording and ends when the patient blinks again. The AI algorithm automatically analyzes the tear film break-up time by red lines that outline the break-up area (Figure 1C). The device reports the first-NIBUT (F-NIBUT) and average-NIBUT (A- NIBUT), which are expressed in s. F-NIBUT and A-NIBUT are defined as the time when the first and half of the rings reflected on the ocular surface using the Placido disc present distortion, respectively. During NIBUT analysis, the AI algorithm also judges tear film stability to determine whether there is a suitable picture to assess tear volume. If there is an appropriate picture, the AI algorithm automatically analyzes the following parameters: tear meniscus height (TMH) and tear meniscus area (TMA), which are expressed in mm and mm2, respectively. TMH was automatically measured at the center of the lower eyelid (Figure 1A), while TMA was automatically measured along the lower eyelid, from the nasal to the temporal area (Figure 1B). Ocular hyperemia Conjunctival and ciliary hyperemia was evaluated through complete picture of the ocular surface focused on the bulbar conjunctiva. This picture was automatically analyzed by the AI identification system, which is able to differentiate between conjunctival and ciliary vessels, generating a conjunctival and ciliary hyperemia map (Figure 1D). This device automatically reports the following nasal and temporal hyperemia parameters: nasal ciliary hyperemia (NCIH), temporal ciliary hyperemia (TCIH), nasal conjunctival hyperemia (NCOH) and temporal conjunctival hyperemia (TCOH), which are expressed as percentages ranging from 0% (no hyperemia) to 100% (the highest level of hyperemia). For this measurement, no Placido disc was placed on the device. Meibomian gland analysis Meibomian gland analysis was performed through focused pictures of the upper and lower meibomian glands, which was taken with the device's infrared non-contact camera after everting the eyelids with a cotton swab. These pictures were automatically analyzed by the AI identification system, reporting the following Meibomian glands loss parameters: upper loss area Meibomian gland (U-LAMG) and lower loss area Meibomian gland (L-LAMG), which are expressed as percentages ranging from 0% (no glandular dropout) to 100% (the highest level of glandular dropout) (Figure 1E and F). In addition, this device also automatically provides a MGD classification based on LAMG, reporting the following parameters: upper Meibomian gland dysfunction grade (U-MGD grade) and lower Meibomian gland dysfunction grade (L- MGD grade), with grades from 1 to 3, where grade 1 indicates LAMG < 30%; grade 2, LAMG between 30% - 60% and grade 3, LAMG > 60%. For this measurement, no Placido disc was placed on the device. Measurements were performed to best preserve the integrity of the tear film to avoid affecting the test results. Tests were carried out in the following sequence: (1) Tear film stability and volume: (1.1) F-NIBUT, (1.2) A-NIBUT, (1.3) TMH and (1.4) TMA; (2) ocular hyperemia: (2.1) NCIH, (2.2) TCIH, (2.3) NCOH, (2.4) TCOH and (3) finally Meibomian gland analysis: (3.1) U-LAMG, (3.2) L-LAMG, (3.3) U-MGD grade and (3.4) L-MGD grade. All measurements were performed independently three times by an experienced and trained optometrist (ABS) on the same day. To avoid errors in the measurements, the pictures were obtained with minimal areas of glare. In addition, the S390L Firefly WDR slit lamp was always located in the same examination room during the study to control fluctuations in temperature and airflow. Statistical analysis Statistical analyses were performed with SPSS statistics software, version 26.0 (IBM Corporation, Armonk, NY, USA). A sample size of 43 patients was estimated using the GRANMO calculator, version 7.12 (Municipal Institute of Medical Research, Barcelona, Spain). It was based on a statistically significant paired difference at 95% confidence and with 80% power of 9.55 ± 4.11 seconds in NIBUT based on previous studies, which used different devices with similar methods to measure NIBUT as S390L Firefly WDR slit lamp.20–22 Prior to the analyses, one eye was randomly selected. The randomization scheme was generated using an online randomizer program ⟨http://www.randomization.com⟩. The Kolmogorov- Smirnov test was used to test the normality of the analyzed variables. Repeated-measures ANOVA (normal distribution) or Friedman's test (non-normal distribution) was performed to check on repeated measures integrity. A post hoc analysis by Wilcoxon’s test with Bonferroni’s adjustment was performed to determine statistically significant differences between pairs of measures. Intra-observer repeatability was assessed using the within-subject standard deviation (Sw), the mean coefficient of variation (CoV) and Bland-Altman plots. Sw was calculated by taking the root mean square error (RMSE) and it was only reported in those variables that had a normal distribution. Using Sw, two values were estimated: (1) the difference between a determined measurement and the true value defined as “1.96 x Sw” and (2) the difference between three repeated measures on the same subject defined as “2.77 x Sw”. Mean CoV was obtained for variables that had a normal and non-normal distribution. Bland-Altman plots were performed for the comparison between pairs of measures. Continuous variables were displayed as the mean ± standard deviation (SD) with interquartile ranges [IQRs], while ordinal categorical variables were expressed as frequencies (n) and percentages (%). The level of significance was P < 0.05 for all of the comparisons. RESULTS Patient characteristics are listed in Table 1. Fifty eyes of 50 patients, 15 (30%) men and 35 (70%) women with a mean age of 55.06 ± 12.96 [30 to 85] years old, were enrolled in the study. Regarding DED related characteristics, the mean OSDI was 43.59 ± 20.67 [10.41 to 90] points. F-NIBUT and A-NIBUT reported a mean value of 6.01 ± 5.94 [0.10 to 19.74] s and 8.34 ± 4.98 [0.07 to 16.73] s, respectively. U-LAMG, U- MGD grade, L-LAMG and L-MGD grade showed a mean value of 41.45 ± 20.85 [14 to 89] %, 1.73 ± 0.64 [1 to 3], 54.5 ± 19.47 [25 to 92] % and 2.15 ± 0.72 [1 to 3], respectively. In relation to patient health, 17 (34%) patients reported systemic diseases, 10 (20%) presented ocular diseases and 34 (68%) declared taking some type of medication during the study. The assessment of the intra-observer repeatability is shown in Table 2. Tear film stability and volume Regarding tear film stability, F-NIBUT (P = 0.723) and A-NIBUT (P = 0.527) presented no statistically significant differences in the repeated measures analysis, with CoVs of 0.48 ± 0.41 [0.02 to 1.00] and 0.34 ± 0.25 [0.02 to 1.00], respectively. A-NIBUT obtained a Sw of 2.78 s. Using this value, the difference between A-NIBUT and the true value will be lower than 5.44 s (1.96 x Sw), while the difference between three A-NIBUT repeated measures on the same subject will be lower than 7.70 s (2.77 x Sw). Bland-Altman plots for F-NIBUT and A-NIBUT are presented in Figure 2. Regarding tear volume, TMH (P = 0.246) and TMA (P = 0.973) showed no statistically significant differences, with CoVs of 0.10 ± 0.05 [0.03 to 0.28] and 0.09 ± 0.06 [0.01 to 0.30], respectively. TMA obtained a Sw value of 0.69 mm2. Using this value, the difference between TMA and the true value will be lower than 0.42 mm2 (1.96 x Sw), while the difference between three TMA repeated measures on the same subject will be lower than 0.61 mm2 (2.77 x Sw) Please refer to supplemental Figure of Bland-Altman plots for TMH and TMA in Supplemental Digital Content 1. Ocular hyperemia NCIH (P = 0.534), TCIH (P = 0.230), NCOH (P = 0.248) and TCOH (P = 0.392) showed no statistically significant differences in the repeated measures analysis, with CoVs of 0.18 ± 0.12 [0.05 to 0.67], 0.18 ± 0.16 [0.00 to 0.75], 0.11 ± 0.10 [0.00 to 0.50] and 0.14 ± 0.11 [0.00 to 0.50], respectively. NCOH obtained a Sw value < 0.001. Using this value, the difference between NCOH and the true value, as well as the differences between three NCOH repeated measures on the same subject will be < 0.001. Please refer to supplemental Figure of Bland-Altman plots for NCIH, TCIH, NCOH and TCOH in Supplemental Digital Content 2. Meibomian gland analysis L-MGD grade presented statistically significant differences in the repeated measures analysis (P = 0.045). A post hoc analysis revealed that significant differences were only present between the first and third measures (P = 0.014), but the CoV was 0.05 ± 0.10 [0.00 to 0.43]. However, U-LAMG (P = 0.937), U- MGD grade (P = 0.449) and L-LAMG (P = 0.926) showed no statistically significant differences, with CoVs of 0.08 ± 0.07 [0.00 to 0.29], 0.04 ± 0.11 [0.00 to 0.43] and 0.05 ± 0.05 [0.01 to 0.28], respectively. L-LAMG obtained a Sw value of 4.11%. Using this value, the estimated difference regarding the true value 20. Bandlitz S, Peter B, Pflugi T, et al. Agreement and repeatability of four different devices to measure non-invasive tear breakup time (NIBUT). Cont Lens Anterior Eye. 2020;43(5):507-511. doi:10.1016/J.CLAE.2020.02.018 21. Molina-Martín A, de Fez D, Piñero DP. Repeatability of non-invasive break-up time measures with a new automated dry eye platform in healthy eyes. Int Ophthalmol. 2020;40(11):2855-2864. doi:10.1007/S10792-020-01470-W 22. Ward CD, Murchison CE, Petroll WM, Robertson DM. Evaluation of the Repeatability of the LacryDiag Ocular Surface Analyzer for Assessment of the Meibomian Glands and Tear Film. Transl Vis Sci Technol. 2021;10(9). doi:10.1167/TVST.10.9.1 23. García-Marqués JV, Martínez-Albert N, Talens-Estarelles C, García-Lázaro S, Cerviño A. Repeatability of Non-invasive Keratograph Break-Up Time measurements obtained using Oculus Keratograph 5M. Int Ophthalmol. 2021;41(7):2473-2483. doi:10.1007/S10792-021-01802-4 24. Hao Y, Tian L, Cao K, Jie Y. Repeatability and Reproducibility of SMTube Measurement in Dry Eye Disease Patients. J Ophthalmol. 2021;2021. doi:10.1155/2021/1589378 25. Yeh TN, Lin MC. Repeatability of Meibomian Gland Contrast, a Potential Indicator of Meibomian Gland Function. Cornea. 2019;38(2):256-261. doi:10.1097/ICO.0000000000001818 26. Schiffman RM, Christianson MD, Jacobsen G, Hirsch JD, Reis BL. Reliability and validity of the Ocular Surface Disease Index. Arch Ophthalmol (Chicago, Ill 1960). 2000;118(5):615-621. doi:10.1001/ARCHOPHT.118.5.615 27. Dougherty BE, Nichols JJ, Nichols KK. Rasch analysis of the Ocular Surface Disease Index (OSDI). Invest Ophthalmol Vis Sci. 2011;52(12):8630-8635. doi:10.1167/IOVS.11-8027 28. Tomlinson A, Bron AJ, Korb DR, et al. The international workshop on meibomian gland dysfunction: report of the diagnosis subcommittee. Invest Ophthalmol Vis Sci. 2011;52(4):2006- 2049. doi:10.1167/IOVS.10-6997F 29. Willcox MDP, Argüeso P, Georgiev GA, et al. TFOS DEWS II Tear Film Report. Ocul Surf. 2017;15(3):366-403. doi:10.1016/J.JTOS.2017.03.006 30. Paulsen AJ, Cruickshanks KJ, Fischer ME, et al. Dry eye in the beaver dam offspring study: prevalence, risk factors, and health-related quality of life. Am J Ophthalmol. 2014;157(4):799- 806. doi:10.1016/J.AJO.2013.12.023 31. Rohit A, Willcox MDP, Brown SHJ, Mitchell TW, Stapleton F. Clinical and biochemical tear lipid parameters in contact lens wearers. Optom Vis Sci. 2014;91(12):1384-1390. doi:10.1097/OPX.0000000000000420 32. Tan LL, Morgan P, Cai ZQ, Straughan RA. Prevalence of and risk factors for symptomatic dry eye disease in Singapore. Clin Exp Optom. 2015;98(1):45-53. doi:10.1111/CXO.12210 33. Mathers WD, Stovall D, Lane JA, Zimmerman MB, Johnson S. Menopause and tear function: the influence of prolactin and sex hormones on human tear production. Cornea. 1998;17(4):353-358. doi:10.1097/00003226-199807000-00002 34. Sullivan DA, Rocha EM, Aragona P, et al. TFOS DEWS II Sex, Gender, and Hormones Report. Ocul Surf. 2017;15(3):284-333. doi:10.1016/J.JTOS.2017.04.001 35. Grasso A, Di Zazzo A, Giannaccare G, et al. Sex Hormones Related Ocular Dryness in Breast Cancer Women. J Clin Med. 2021;10(12). doi:10.3390/JCM10122620 36. Kojima T, Ibrahim OMA, Wakamatsu T, et al. The impact of contact lens wear and visual display terminal work on ocular surface and tear functions in office workers. Am J Ophthalmol. 2011;152(6). doi:10.1016/J.AJO.2011.05.025 37. Abusharha AA, Pearce EI. The effect of low humidity on the human tear film. Cornea. 2013;32(4):429-434. doi:10.1097/ICO.0B013E31826671AB 38. López-Miguel A, Tesón M, Martín-Montañez V, et al. Dry eye exacerbation in patients exposed to desiccating stress under controlled environmental conditions. Am J Ophthalmol. 2014;157(4). doi:10.1016/J.AJO.2014.01.001 39. Torricelli AAM, Novaes P, Matsuda M, Alves MR, Monteiro MLR. Ocular surface adverse effects of ambient levels of air pollution. Arq Bras Oftalmol. 2011;74(5):377-381. doi:10.1590/S0004-27492011000500016 40. Torricelli AAM, Matsuda M, Novaes P, et al. Effects of ambient levels of traffic-derived air pollution on the ocular surface: analysis of symptoms, conjunctival goblet cell count and mucin 5AC gene expression. Environ Res. 2014;131:59-63. doi:10.1016/J.ENVRES.2014.02.014 41. Abusharha AA, Pearce EI, Fagehi R. Effect of Ambient Temperature on the Human Tear Film. Eye Contact Lens. 2016;42(5):308-312. doi:10.1097/ICL.0000000000000210 42. Niedernolte B, Trunk L, Wolffsohn JS, Pult H, Bandlitz S. Evaluation of tear meniscus height using different clinical methods. Clin Exp Optom. 2021;104(5):583-588. doi:10.1080/08164622.2021.1878854 43. Nichols JJ, Nichols KK, Puent B, Saracino M, Mitchell GL. Evaluation of tear film interference patterns and measures of tear break-up time. Optom Vis Sci. 2002;79(6):363-369. doi:10.1097/00006324-200206000-00009 44. Rodriguez JD, Johnston PR, Ousler GW, Smith LM, Abelson MB. Automated grading system for evaluation of ocular redness associated with dry eye. Clin Ophthalmol. 2013;7:1197. doi:10.2147/OPTH.S39703 45. García-Montero M, Rico-Del-Viejo L, Lorente-Velázquez A, Martínez-Alberquilla I, Hernández- Verdejo JL, Madrid-Costa D. Repeatability of Noninvasive Keratograph 5M Measurements Associated With Contact Lens Wear. Eye Contact Lens. 2019;45(6):377-381. doi:10.1097/ICL.0000000000000596 46. Nichols KK, Foulks GN, Bron AJ, et al. The international workshop on meibomian gland dysfunction: executive summary. Invest Ophthalmol Vis Sci. 2011;52(4):1922-1929. doi:10.1167/IOVS.10-6997A 47. Arita R. Meibography: A Japanese Perspective. Invest Ophthalmol Vis Sci. 2018;59(14):DES48- DES55. doi:10.1167/IOVS.17-23631 48. Maskin SL, Testa WR. Infrared video meibography of lower lid meibomian glands shows easily distorted glands: Implications for longitudinal assessment of atrophy or growth using lower lid meibography. Cornea. 2018;37(10):1279-1286. doi:10.1097/ICO.0000000000001710 49. Bron AJ, de Paiva CS, Chauhan SK, et al. TFOS DEWS II pathophysiology report. Ocul Surf. 2017;15(3):438-510. doi:10.1016/J.JTOS.2017.05.011 50. Setu MAK, Horstmann J, Schmidt S, Stern ME, Steven P. Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography. Sci Rep. 2021;11(1). doi:10.1038/s41598-021-87314-8 FIGURE LEGENDS FIG. 1 Objective and non-invasive tests analyzed automatically by the S390L Firefly WDR slit lamp: (A) TMH; (B) TMA; (C) Tear film stability; (D) Ocular hyperemia; (E) U-LAMG and (F) L-LAMG. FIG. 2 Bland-Altman plots for tear film stability repeated measures: (A), (C) and (E) refers to F-NIBUT. (B), (D) and (F) refer to A-NIBUT. The 95 % confidence intervals and mean differences are shown as a twodash and solid lines, respectively. FIG. 3 Bland-Altman plots for meibomian gland repeated measures: (A), (C) and (E) refer to U-LAMG. (B), (D) and (F) refer to L-LAMG. The 95 % confidence intervals and mean differences are shown as a twodash and solid lines, respectively. SUPPLEMENTAL DIGITAL CONTENT LEGEND Supplemental Digital Content 1. Bland-Altman plots for meniscometry repeated measures: (A), (C) and (E) refer to TMH. (B), (D) and (F) refer to TMA. The 95 % confidence intervals and mean differences are shown as a twodash and solid lines, respectively. Supplemental Digital Content 2. Bland-Altman plots for ocular hyperemia repeated measures: (A), (E) and (I) refer to NCIH. (B), (F) and (J) refer to TCIH; (C), (G) and (K) refer to NCOH. (D), (H) and (L) refer to TCOH. The 95 % confidence intervals and mean differences are shown as a twodash and solid lines, respectively. TABLE 1. Characteristics of dry eye patients Characteristics n = 50 Demographics Age (years), mean (SD) 55.06 (12.96) Sex, male/female (%) 15 (30) / 35 (70) Race, Caucasian (%) 50 (100) Contact lens wearer, n (%)a 22 (44) Related to DED OSDI, mean (SD)b 43.59 (20.67) F-NIBUT (s), mean (SD) 6.01 (5.94) A-NIBUT (s), mean (SD) 8.34 (4.98) U-LAMG (%), mean (SD) 41.45 (20.85) U-MGD grade, mean (SD)c 1.73 (0.64) L-LAMG (%), mean (SD) 54.53 (19.47) L-MGD grade, mean (SD)c 2.15 (0.72) Systemic diseases, n (%) Thyroid disease 7 (14) Rheumatoid arthritis 5 (10) Rosacea 3 (6) Diabetes mellitus 2 (4) Sjögren's syndrome 2 (4) Atopy 1 (2) Ocular diseases, n (%) Cataracts 8 (16) Glaucoma 5 (10) AMD 3 (6) Diabetic retinopathy 1 (2) Medications, n (%) Antidepressants 14 (28) Anxiolytics 14 (28) Arterial Hypertension drugs 13 (26) Hormonal Therapy 10 (20) Isotretinoin 5 (10) Antihistamines 5 (10) Topical hypotensive drugs 5 (10) Antineoplastic drugs 2 (4) Diuretics 2 (4) Anti-VEGF 2 (4) Ocular surgery, n (%) Cataract surgery 8 (16) AMD, age-related macular degeneration; Anti-VEGF, anti-vascular endothelial growth factor therapy; A-NIBUT, average non-invasive break-up time; F-NIBUT, first non-invasive break-up time; L-LAMG, lower loss area meibomian gland; L-MGD grade, lower meibomian gland dysfunction grade; OSDI, ocular surface disease index; U-LAMG, upper loss area meibomian gland; U-MGD Grade, upper meibomian gland dysfunction grade. a Wearing contact lenses for more than two years. b Values from 0 to 100. c Values from 1 to 3. TABLE 2. Evaluation of the intra-observer repeatability Variables Measuresd Coefficient of variationd P-values First Second Third Average F-NIBUT (s) 6.70 ± 6.79 [0.04 to 20.47] 6.68 ± 6.85 [0.04 to 20.67] 6.55 ± 6.51 [0.04 to 22.32] 6.64 ± 6.71 [0.04 to 22.32] 0.48 ± 0.41 [0.02 to 1.00] P = 0.723b A-NIBUT (s) 8.88 ± 6.04 [0.55 to 20.47] 8.96 ± 6 [0.33 to 21.00] 9.05 ± 5.64 [0.30 to 22.32] 8.96 ± 5.89 [0.30 to 22.32] 0.34 ± 0.25 [0.02 to 1.00] P = 0.958c TMH (mm) 0.21 ± 0.06 [0.10 to 0.48] 0.21 ± 0.06 [0.12 to 0.48] 0.20 ± 0.06 [0.12 to 0.42] 0.21 ± 0.06 [0.10 to 0.48] 0.10 ± 0.05 [0.03 to 0.28] P = 0.246b TMA (mm2) 2.16 ± 0.79 [0.56 to 4.48] 2.15 ± 0.77 [0.56 to 4.15] 2.16 ± 0.78 [0.82 to 4.47] 2.16 ± 0.78 [0.56 to 4.48] 0.09 ± 0.06 [0.01 to 0.30] P = 0.973c U-LAMG (%) 43.48 ± 19.29 [6 to 95] 43.54 ± 19.25 [10 to 94] 43.38 ± 19.84 [11 to 97] 43.47 ± 19.46 [6 to 97] 0.08 ± 0.07 [0.00 to 0.29] P = 0.937b U-MGD Gradea 1.86 ± 0.70 [1 to 3] 1.82 ± 0.69 [1 to 3] 1.82 ± 0.72 [1 to 3] 1.83 ± 0.70 [1 to 3] 0.04 ± 0.11 [0.00 to 0.43] P = 0.449b L-LAMG (%) 67.26 ± 13.93 [42 to 99] 67.38 ± 15.69 [26 to 100] 67.06 ± 15.31 [30 to 99] 67.23 ± 14.98 [26 to 100] 0.05 ± 0.05 [0.01 to 0.28] P = 0.926c L-MGD Gradea 2.54 ± 0.50 [2 to 3] 2.52 ± 0.54 [1 to 3] 2.42 ± 0.53 [1 to 3] 2.49 ± 0.52 [1 to 3] 0.05 ± 0.10 [0.00 to 0.43] P = 0.045b,* NCIH (%) 7.34 ± 3.64 [2 to 20] 7.44 ± 3.76 [2 to 19] 7.29 ± 3.52 [1 to 18] 7.36 ± 3.64 [1 to 20] 0.18 ± 0.12 [0.05 to 0.67] P = 0.556b TCIH (%) 8.89 ± 3.56 [1 to 18] 8.70 ± 3.68 [1 to 16] 8.76 ± 3.75 [1 to 15] 8.78 ± 3.66 [1 to 18] 0.18 ± 0.16 [0.00 to 0.75] P = 0.222b NCOH (%) 9.2 ± 2.22 [3 to 18] 9.14 ± 2.91 [4 to 18] 9.22 ± 3.01 [2 to 18] 9.19 ± 2.71 [2 to 18] 0.11 ± 0.10 [0.00 to 0.50] P = 0.245c TCOH (%) 10.11 ± 3.34 [4 to 23] 10.04 ± 3.21 [5 to 25] 10.19 ± 3.45 [4 to 16] 10.11 ± 3.33 [4 to 25] 0.14 ± 0.11 [0.00 to 0.50] P = 0.383b A-NIBUT, average non-invasive break-up time; F-NIBUT, first non-invasive break-up time; L-LAMG, lower loss area meibomian gland; L-MGD grade, lower meibomian gland dysfunction grade; NCIH, nasal ciliary hyperemia; NCOH, nasal conjunctival hyperemia; TCIH, temporal ciliary hyperemia; TCOH, temporal conjunctival hyperemia; TMA, tear meniscus area; TMH, tear meniscus height; U-LAMG, upper loss area meibomian gland; U-MGD Grade, upper meibomian gland dysfunction grade. a Values from 1 to 3. b Nonparametric repeated measures ANOVA with Friedman’s Test. c Parametric repeated measures ANOVA. d Expressed as Mean ± SD [IQR]. * Level of significance P < 0.050.