Purpose. To establish the relationships between commonly used questionnaires including Dry Eye Questionnaire, McMonnies Questionnaire, and Ocular Surface Disease Index, and to test the construct and face validity of the simple Subjective Evaluation of Symptom of Dryness.
Methods. Ninety-seven non-contact lens wearing subjects were enrolled in the study and classified into either a “dry” and “non-dry” group using a single score from an initially applied subjective evaluation of symptom of dryness. The four questionnaires were then completed in a random order. The unidimensionality and accuracy of the responses was assessed using Rasch and receiver (or relative) operating characteristics curve analysis and the characteristics of and association between symptoms were compared using non-parametric statistics.
Results. The responses from the Dry Eye Questionnaire, McMonnies Questionnaire, and Ocular Surface Disease Index met the Rasch analysis criterion of unidimensionality. Each test separated the symptomatic and asymptomatic groups well [all receiver (or relative) operating characteristics area-under-the-curve statistics at least 0.88] and there were significant associations between the results from each questionnaire (all Spearman ρ at least 0.64).
Conclusions. The results illustrate that different questionnaire-based instruments examining symptoms in controls and symptomatic subjects derive unidimensional data that are similar inasmuch as the overall scores are highly correlated. The data also point to the utility of a quick, three-question screening tool in dry eye research.
Centre for Contact Lens Research, School of Optometry, University of Waterloo, Waterloo, Ontario, Canada
Received September 23, 2007; accepted May 1, 2008.