Purpose

Undiagnosed or inadequately treated dry eye disease (DED) decreases the quality of life. We aimed to investigate the reliability, validity, and feasibility of the DryEyeRhythm smartphone application (app) for the diagnosis assistance of DED.

Methods

This prospective, cross-sectional, observational, single-center study recruited 82 participants (42 with DED) aged ≥20 years (July 2020–May 2021). Patients with a history of eyelid disorder, ptosis, mental disease, Parkinson's disease, or any other disease affecting blinking were excluded. Participants underwent DED examinations, including the Japanese version of the Ocular Surface Disease Index (J-OSDI) and maximum blink interval (MBI). We analyzed their app-based J-OSDI and MBI results. Internal consistency reliability and concurrent validity were evaluated using Cronbach's alpha coefficients and Pearson's test, respectively. The discriminant validity of the app-based DED diagnosis was assessed by comparing the results of the clinical-based J-OSDI and MBI. The app feasibility and screening performance were evaluated using the precision rate and receiver operating characteristic curve analysis.

Results

The app-based J-OSDI showed good internal consistency (Cronbach's α = 0.874). The app-based J-OSDI and MBI were positively correlated with their clinical-based counterparts (r = 0.891 and r = 0.329, respectively). Discriminant validity of the app-based J-OSDI and MBI yielded significantly higher total scores for the DED cohort (8.6 ± 9.3 vs. 28.4 ± 14.9, P < 0.001; 19.0 ± 11.1 vs. 13.2 ± 9.3, P < 0.001). The app's positive and negative predictive values were 91.3% and 69.1%, respectively. The area under the curve (95% confidence interval) was 0.910 (0.846–0.973) with concurrent use of the app-based J-OSDI and MBI.

Conclusions

DryEyeRhythm app is a novel, non-invasive, reliable, and valid instrument for assessing DED.

中文翻译：

---

DryEyeRhythm：一款可靠有效的干眼诊断辅助智能手机应用程序

目的

未确诊或治疗不当的干眼病 (DED) 会降低生活质量。我们旨在调查 DryEyeRhythm 智能手机应用程序 (app) 用于 DED 诊断辅助的可靠性、有效性和可行性。

方法

这项前瞻性、横断面、观察性、单中心研究招募了 82 名参与者（42 名患有 DED），年龄≥20 岁（2020 年 7 月至 2021 年 5 月）。有眼睑疾病、上睑下垂、精神疾病、帕金森病或任何其他影响眨眼的疾病史的患者被排除在外。参与者接受了 DED 检查，包括日本版的眼表疾病指数 (J-OSDI) 和最大眨眼间隔 (MBI)。我们分析了他们基于应用程序的 J-OSDI 和 MBI 结果。内部一致性信度和同时效度分别使用 Cronbach α 系数和 Pearson 检验进行评估。通过比较基于临床的 J-OSDI 和 MBI 的结果来评估基于应用程序的 DED 诊断的判别有效性。

结果

基于应用程序的 J-OSDI 显示出良好的内部一致性 (Cronbach's α = 0.874)。基于应用的 J-OSDI 和 MBI 与其基于临床的对应物呈正相关（分别为 r = 0.891 和 r = 0.329）。基于应用程序的 J-OSDI 和 MBI 的判别效度显着提高了 DED 队列的总分（8.6 ± 9.3 对 28.4 ± 14.9，P  < 0.001；19.0 ± 11.1 对 13.2 ± 9.3，P  < 0.001）。该应用程序的阳性和阴性预测值分别为 91.3% 和 69.1%。曲线下面积（95% 置信区间）为 0.910（0.846-0.973），同时使用基于应用程序的 J-OSDI 和 MBI。

结论

DryEyeRhythm 应用程序是一种新颖、无创、可靠且有效的 DED 评估工具。