Abstract
OBJECTIVES: To determine the efficacy of widely available subtype clinical tests to characterize evaporative dry eye disease (EDED) related to meibomian gland dysfunction (MGD) compared to normal and to validate those clinical cut points in an independent sample. METHODS: A diagnostic accuracy study (52 subjects), an investigator-masked study, was followed by a larger independent sample (364 subjects) analysis to confirm efficacy in normal and EDED subjects. All subjects were 18 years of age and older and were classified using a battery of clinical tests for dry eye that included symptoms, tear meniscus height, tear stability, ocular staining, evaporative-specific tests, and the Schirmer I test. RESULTS: Normal (nondry eye; n = 26) and EDED (n = 26) subjects completed the efficacy study. The global tests of tear breakup time, staining, and symptoms all produced AUCs ≥ 0.70, representing acceptable discrimination. EDED-specific tests of eyelid marginal signs, gland secretion quality, and gland loss did not demonstrate acceptable test efficacy or differences between normal and EDED subjects. In a larger, independent sample of normal and EDED subjects, gland secretion quality and eyelid marginal score achieved acceptable diagnostic levels: AUCs of 0.789 (CI: 0.734-0.844) and 0.729 (CI: 0.648-0.810), respectively, but not lipid interferometry grade or lower eyelid gland dropout estimated using meiboscopy. CONCLUSIONS: Meibomian gland secretion quality is an efficient and useful functional indicator in EDED and should be incorporated into core outcome sets for this dry eye subtype.