Objective: To assess the performance in the detection of cases of rheumatoid arthritis (RA) and the spondyloarthropathies (SpA) of a questionnaire suitable for use in telephone surveys conducted by patient interviewers.
Methods: A questionnaire was designed with reference to the signs, symptoms, and epidemiological criteria for RA (ACR 1987) and SpA (ESSG 1991). Three groups of respondents were recruited from the rheumatology outpatient clinics of 10 university hospitals: 235 with RA, 175 with SpA, and 195 controls with other rheumatological disorders. All diagnoses were confirmed by a rheumatologist. Patient from self help groups and social organisations were trained by a polling company professional to conduct a standard telephone interview using the new questionnaire.
Results: In an RA-control comparison, logistic regression showed that a set of five items, predominantly ACR criteria, were the most informative. Self reported diagnosis performed best (sensitivity 0.99, specificity 0.87). In an SpA-control comparison, a set of three items from the ESSG criteria were the most informative, with self reported diagnosis again performing best (sensitivity 0.85, specificity 0.96). Overall agreements with clinical diagnoses were 97.7% for RA and 94.4% SpA, dropping to 90.4% and 79.1%, respectively, when self reported diagnosis was excluded. Without self reported diagnosis, questions about peripheral joint and spinal pain made significant contributions to diagnostic performance.
Conclusion: A questionnaire in plain language was developed for use in detecting cases of RA and SpA. It performed satisfactorily when administered by patient interviewers and is now available for epidemiological surveys of the general population.