Purchase this article with an account.
R.L. Chalmers, C. Begley; Repeatability of Habitual Symptoms as Measured by the Dry Eye Questionnaire (DEQ) . Invest. Ophthalmol. Vis. Sci. 2004;45(13):3455.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To measure the repeatability of ocular surface symptoms in the Dry Eye Questionnaire (DEQ) and estimate a clinically significant change in symptoms for use in clinical trials. Methods: Normals (48) and dry eye patients (162) were recruited for an observational study of ocular signs and symptoms. Subjects completed the DEQ twice at a two week interval with no change in dry eye treatment or management. At the second visit, they also made a global assessment of change in their dry eye condition. The item agreement was compared with Cohen’s Kappa for those who reported no change (NC) in their dry eye condition (n=167) and those 39 who reported change (C). The estimate of a clinically significant change results from examining the responsiveness of various questions. Results: The range of Kappa for frequency questions was .293 (Tired eyes) to .594 (Dryness) for the NC patients and .326 (Burning) to .412 (Dryness) for the C subjects, where Kappas were lower in most cases. PM Intensity questions were most responsive, with Kappa values from .361 (PM Tired Eyes) to .472 (PM Blurry Vision) for NC subjects and from .133 (PM Dryness) to .181 (PM Tired Eyes) among C subjects (p<0.05). AM symptoms showed no significant difference in repeatability across groups. Frequency answers were consistent within 1 unit on the scale 95–98% of the time for NC subjects and 82–95% for C subjects. PM intensity questions were answered consistently within 1 unit 86–95% for NC subjects and 67–87% for C subjects, indicating that PM discomfort, grittiness/scratchiness, tired eyes and changeable, blurry vision were most responsive to the subjects’ global report of change in dry eye. Conclusions: This DEQ data shows a high degree of consistency in habitual symptoms among patients who reported no change and without treatment intervention at an interval of two weeks. The most responsive questions were PM intensity questions, reflecting change among those who reported global change. We propose that a clinically significant treatment effect for use in clinical trials would be > 1 unit on the intensity scale.
This PDF is available to Subscribers Only