Lid wiper epitheliopathy in patients with dry eye refractory to conventional medical treatment

Jungah Huh, In Ho Woo, Youngsub Eom, Su Yeon Kang, Hyo Myung Kim, Jong Suk Song

    Research output: Contribution to journalArticlepeer-review


    Purpose: We evaluated the effects of lid wiper epitheliopathy on the clinical manifestations of dry eye in patients refractory to conventional medical treatment. Methods: Forty-six patients (46 eyes) completed the subjective Ocular Surface Disease Index (OSDI), and we obtained objective measures including the tear break-up time (TBUT), the National Eye Institute (NEI) corneal staining score, tear osmolarity, and lid wiper epitheliopathy as revealed on photographs taken using a yellow filter after fluorescein instillation. The images were graded using the Korb B protocol. Results: The mean OSDI score was 48.06 ± 21.19; 34 patients (73.9%) were had scores ≥33. Lid wiper epitheliopathy was evident in 41 (89.1%), and the epitheliopathy grade and OSDI score were correlated (r = 0.56, p < 0.01). The NEI score was also positively correlated with the OSDI score (r = 0.54, p < 0.01), but the mean value was low (1.59 ± 2.13). The OSDI score did not correlate significantly with either the TBUT or tear osmolarity (r = -0.16, p = 0.279; r = 0.16, p = 0.298, respectively). Conclusions: Lid wiper epitheliopathy was present in about 90% of patients with dry eye who visited our tertiary hospital, and a significant correlation was evident between the epitheliopathy grade and the OSDI score.

    Original languageEnglish
    Pages (from-to)718-723
    Number of pages6
    JournalJournal of Korean Ophthalmological Society
    Issue number8
    Publication statusPublished - 2018 Aug


    • Correlation
    • Dry eye
    • Lid wiper epitheliopathy
    • Medical
    • Refractory

    ASJC Scopus subject areas

    • Ophthalmology


    Dive into the research topics of 'Lid wiper epitheliopathy in patients with dry eye refractory to conventional medical treatment'. Together they form a unique fingerprint.

    Cite this