The prevalence of psoriatic arthritis in people with psoriasis

Abstract

To determine the prevalence of psoriatic arthritis (PsA) using the ClASsification criteria for Psoriatic ARthritis (CASPAR) for classification. People with psoriasis were identified from the computerized morbidity indices of 2 large UK general practices, total population 22,500. Questionnaires were mailed to all 633 patients thus identified. Of the respondents, a 50% sample was assessed clinically and a proportion had blood samples and radiographs taken. Patients labeled as having psoriasis were also cross-referenced with a local secondary care morbidity index for PsA and rheumatoid arthritis. Figures for the prevalence of PsA were estimated from these data. One hundred sixty-eight questionnaires were returned (response rate 27%) and 93 people (55% of questionnaire respondents) were examined. Of these 93 people, 12 (4 of whom were cross-referenced to the hospital database) were thought to have PsA clinically, all fulfilling the CASPAR criteria for PsA. Six of the 93 examined patients did not have psoriasis or a family history of psoriasis and had no historical features or clinical signs of psoriasis on interview and examination. Extrapolating from the data of those people actually examined, the estimated (corrected) prevalence was 13.8% (95% confidence interval 7.1-24.1%). The estimated prevalence of PsA in this population, using the CASPAR criteria, was 13.8%. Misclassification of psoriasis and arthritis, and response bias, indicate that this is probably an overestimate.