The European Community Study Group on diagnostic criteria for Sjögren's syndrome. Sensitivity and specificity of tests for ocular and oral involvement in Sjögren's syndrome.
AIM--To establish a definitive set of diagnostic criteria in a multicentre European study a selected number of oral and ocular tests were performed on a large number of patients with Sjögrens Syndrome (SS) and controls. The diagnostic accuracy of each test for patients with primary and secondary SS and for controls at different ages, was studied. METHODS--Each centre received a clinical chart describing the series of tests to be conducted. The tests included: questionnaires for dry eye and dry mouth symptoms, Schirmer's-I-test (ScT), tear fluid lactoferrin level (TFLL), break-up time (BUT) and rose Bengal score (RBS) for the eye evaluation; unstimulated and stimulated whole saliva collection (UWSC and SWSC), salivary gland scintigraphy (SGS), parotid sialography (PS) and minor salivary gland biopsy (MSGB) for oral involvement. RESULTS--Data from 22 centres and 11 countries was collected on a total of 447 patients with SS (246 with primary SS and 201 with secondary SS) and 246 controls (of whom 113 had a connective tissue disease without SS). Among the ocular symptoms, the feeling of dry eye and 'sand in the eye' were the ones most commonly recorded in patients with SS. Similarly, the feeling of dry mouth, appearing either spontaneously or when the patient was eating or breathing, was the most frequent subjective oral symptom. Among the ocular tests, ScT showed the best balance between sensitivity and specificity (76.9% and 72.4% respectively), while RBS was the most specific test (81.7%). ScT and RBS gave also sufficiently concordant results. TFLL and BUT gave considerably less reliable results, which were not concordant with each other or with the other ocular tests. The quantitative lacrimal tests ScT and TFLL produced significantly different results in elderly controls, while RBS did not. Abnormal results for all of the ocular tests were less marked and less frequent in patients with secondary SS than in those with primary SS. The oral tests (except SWSC) were generally more reliable than the ocular tests in diagnosing SS. In particular, PS was the most specific diagnostic tools (100%), while MSGB (where the presence of at least one inflammatory focus was considered as indicative for the diagnosis) showed a good balance between sensitivity and specificity (82.4% and 86.2%, respectively). The tests showed a good degree of agreement, and, with the exception of UWSC, were not influenced by age. In the oral, as in the ocular tests, abnormal results were less frequent and less marked in patients with secondary SS. CONCLUSIONS--The results clearly show that ScT and RBS (for the eye evaluation), and SGS, PS, MSGB and UWSC (for salivary gland involvement) are the most reliable tests for the diagnosis of SS. The clinician should be aware, however, that the test results may vary depending on the age of the patient and the type of SS (primary or secondary).