Comparison of screening questionnaires to identifypsoriatic arthritis in a primary-care population: across-sectional study

Background: many questionnaires are available for assessment of psoriatic arthritis (PsA), but there is little evidence comparing them. Objectives: to test the proposed CONTEST questionnaire, which was developed to identify patients with psoriasis who have undiagnosed PsA, and compare it with the validated Psoriasis Epidemiology Screening Tool (PEST) questionnaire in a primary-care setting. Methods: a random sample of adult patients with psoriasis and no diagnosis of arthritis was identified from five general practice surgeries in Yorkshire, U.K. Consenting patients completed both questionnaires and were assessed by a dermatologist and rheumatologist. Diagnosis of PsA was made by the assessing rheumatologist. Receiver operator characteristic (ROC) curve analysis examined the sensitivity and specificity of potential cut points. Results: in total 932 packs were sent to recruit 191 (20·5%) participants. Of these, 169 (88·5%) were confirmed to have current or previous psoriasis. Using physician diagnosis 17 (10·1%) were found to have previously undiagnosed PsA, while 90 (53·3%) had another musculoskeletal complaint and 62 (36·7%) had no musculoskeletal problems. Using ROC curve analysis, all of the questionnaires showed a significant ability to identify PsA. The area under the curve (AUC) for the CONTEST questionnaires was slightly higher than that of PEST (0·69 and 0·70 vs. 0·65), but there was no significant difference identified. Examining the sensitivities and specificities for the different cut points suggested that a PEST score ≥ 2 would perform better in this dataset, and the optimal scores for CONTEST and CONTEST plus joint manikin were 3 and 4, respectively. Conclusions: the accuracy of the questionnaires to identify PsA appeared similar, with a slightly higher AUC for the CONTEST questionnaires. The optimal cut points in this study appeared lower than in previous studies