The Performance of the Contact Lens Dry Eye Questionnaire as a Screening Survey for Contact Lens-related Dry Eye

Abstract

PURPOSE: The contact lens dry eye questionnaire (CLDEQ) is a self-administered survey developed to examine the distribution of dry eye symptoms among contact lens wearers. In this report, we examine the CLDEQ as a screening survey for contact lens-related dry eye and compare it with McMonnies' questionnaire. METHODS: The CLDEQ and McMonnies' questionnaire were administered to 367 unselected contact lens wearers at six clinics across the United States and Canada. After completion of the surveys, doctors unaware of the survey results completed a separate form indicating contact lens-related dry eye diagnosis at the end of a nondirected clinical examination. The CLDEQ is composed of nine habitual symptom subscales and a self-diagnosis question, which were tested for their predictive value for a diagnosis of contact lens-related dry eye. McMonnies' instrument was scored with use of the algorithm suggested in the literature. Sensitivity, specificity, and receiver operator characteristic (ROC) curve analyses were performed for each instrument on the basis of logistic regression results. RESULTS: The area under the ROC curve for the CLDEQ was 0.74, indicating moderate contact lens-related dry eye discrimination, and the Hosmer-Lemeshow goodness-of-fit test indicated that the CLDEQ was well calibrated (p = 0.84). The area under the ROC curve for McMonnies' questionnaire was 0.56, indicating poorer discrimination, and the Hosmer-Lemeshow goodness-of-fit test indicated it was also poorly calibrated (p = 0.08) for contact lens wearers. CONCLUSIONS: These analyses suggest that the CLDEQ is capable of discriminating contact lens-related dry eye and is accurate in doing so, especially in comparison with McMonnies' questionnaire. The CLDEQ is an efficient screening survey and may be used in future clinical research and epidemiologic studies of contact lens-related dry eye.