[Lack of concordance between dry eye syndrome questionnaires and diagnostic tests].

PURPOSE To report the prevalence of dry eye syndrome (DES) in a subset of patients > 50 years old in Valladolid, Spain, calculate internal validity of two DES screening questionnaires, and correlate the results with DES diagnostic tests. METHODS Patients > 50 years-old were randomly selected from the medical network census in Valladolid; they answered the modified McMonnies questionnaire (Q1) and the Ocular Surface Index questionnaire (Q2) and then underwent an ophthalmic evaluation which included: tear break-up time (TBUT), fluorescein and Rose Bengal staining, and Schirmer with anesthesia were performed. Descriptive analysis, test concordance analysis, and Q1 and Q2 reliability using ROC curves were performed. RESULTS A total of 270 subjects (58.2% female; 41.8% male) with an average age of 64.5 years (95% CI: 63.3 - 65.7) were enrolled. The prevalence of DES based on the Schirmer-Tear break-up time tests combination was 24.2%. The mean Q1 score was 1.2, and 1.4 for Q2. Grittiness and discomfort in dry environments (28.9%) were the most common symptoms recorded for Q1. For Q2 "grittiness" (99.2%) and "burning sensation" (98.9%)were the most common symptoms. With the Schirmer-TBUT combination, DES prevalence in Valladolid was 24.2%, 45.8% with TBUT, and 45.6% with Schirmer test. Values for the areas under the ROC curve for Q1, Q2, and most objective tests were < 0.5, except for Q2 and TBUT (0.503). Cronbach alpha was > 0.7 for all Q1 and Q2 items. Statistically significant concordance was reported for Schirmer - TBUT (κ=0.14) and TBUT- fluorescein staining (κ=0.09); (P<0.05). CONCLUSIONS Q1 and Q2 were unreliable for DES diagnosis and showed poor internal consistency. Fluorescein and Rose Bengal staining were the most concordant tests for DES diagnosis. Finally, there was no correlation between either screening questionnaire and objective DES diagnostic tests.