Analysis of Accuracy of Tear Breakup Time (TBUT) and Non-invasive TBUT

Abstract

Purpose: To compare the accuracy of invasive and non-invasive tear breakup time methods Methods: Tear breakup time (TBUT) was measured by using a fluorescein strip (Haag-Streit Strips, Haag-Streit, USA) and non-invasive tear break up time (NIBUT) was measured by using Keratograph 5M (K5M), Wavefront Analyzer (KR-1W), and Auto refraction keratometry (KR-8100P) for the 56 university students (112 eyes) who neither had eye disease nor had undergone refractive surgery. For the dry eyes screening, OSDI (Ocular Surface Disease Index) questionnaire was used. Agreement between tear break up methods was analyzed by using Bland & Altman Plot. Sensitivity, specificity and area under the curve (AUC) and cut-off value were analyzed by using receiver operating characteristic (ROC) curve Results: TBUT was significantly shorter than NIBUT (P<0.05) and NIBUTs measured by KR-1W, KR-8100P, K5M were not significantly different (F=2.69, P=0.06). NIBUTs were shown moderate correlation among them(r=0.36, r=0.44, r=0.35), however TBUT were shown correlation only with NIBUT measured by K5M. In the agreements analysis, there was no difference among NIBUTs, however TBUT and NIBUTs were shown significantly different. Sensitivities and specificities for the dry eye screening of the TBUT and NIBUTs measured by KR-1W, KR-8100P and K5M method were 25% and 80%, 50% and 69%, and 86% and 34%, 31% and 82%, respectively. The sensitivity of KR-8100P measurements was the highest and the specificity of K5M measurements was the highest. Conclusion: NIBUT was more accuracy than TBUT and agreement of measurements among NIBUTs was good, and among the NIBUT measurements the KR-8100P measurements showed high accuaracy that it could be useful for dry eye screening.