Diagnostic Performance of McMonnies Questionnaire as a Screening Survey for Dry Eye: A Multicenter Analysis

Purpose. To evaluate the diagnostic performance of the McMonnies questionnaire as a screening survey for dry eye in Chinese outpatients. Methods. The questionnaire was self-administered by 27,999 patients with dry eye symptoms. A thorough ophthalmic examination including tear break-up time (TBUT), fluorescein staining, and Schirmer I test was completed to make a clinical diagnosis of dry eye. Reliability, validity, and accuracy of the McMonnies questionnaire were assessed. Results. The McMonnies questionnaire showed poor internal consistency (Cronbach α = 0.37), but excellent validity as the scores correlated with TBUT (Spearman test, r = −0.322, P < 0.001) and Schirmer I test (Spearman's test, r = −0.370, P < 0.001), and significantly differed between the dry eye and control groups (2-sample t-test, t = 69.51, P < 0.001). The area under the receiver-operating characteristics (ROC) curve (AUC) was 0.729, suggesting moderate accuracy in identifying dry eye and non-dry eye patients. However, the AUCs varied significantly in different gender and age subgroups (z test, P < 0.001), as the discriminating ability declined with age. Analysis of the ROC curves also revealed that different cut-off points should be employed for each subgroup to achieve the same level of accuracy. Conclusions. The McMonnies questionnaire demonstrates moderate diagnostic value, and different cut-off points should be selected for various study populations.