N. Gouyette, T. Passeron, A. Cantagrel, E. Audureau, E. Dernis, D. Jullien, D. Lons Danic, M. A Richard, F. Roux, F. Lioté, M. Bagot, P. Claudepierre, P. Hilliquin
Background
Dermatologists are recommended to ask psoriasis patients about musculoskeletal complaints to allow early detection and treatment of psoriatic arthritis (PsA). Screening tools have been developed to help identifying patients warranting further rheumatologic assessment, but evidence suggests room for improvement in their diagnostic value and ease of use for outpatient practice.
Objective
To develop and internally validate a brief tool for dermatologists to screen patients to refer to a rheumatologist for PsA diagnosis.
Methods
After literature review, 23 items were selected, covering pain at various locations and inflammatory signs of PsA. The validation study was conducted in medically diagnosed psoriasis patients consecutively recruited between 2012-2014 (Saint Joseph Hospital, Paris, France). Patients were enrolled by a dermatologist who helped to complete the questionnaire. Diagnosis of PsA was established by a rheumatologist based on CASPAR criteria. Multivariate logistic regression models were performed to build the scale, assessing discrimination through sensitivity, specificity and area under the ROC curve (AUC). Final model was internally validated using bootstrapping techniques.
Results
168 patients were recruited, of whom 9 were excluded for known PsA and 21 did not attend the rheumatologist consultation. Out of 137 included patients (median age 43 years, 59.6% men), 21 (15.3%) had a PsA diagnosis. Final regression model retained 4 independent items, including evocative signs of dactylitis, inflammatory heel pain, bilateral buttock pain and peripheral joint pain with swelling in patients aged<50. A total score (the PURE-4) was computed (0-4 points) that demonstrated excellent discriminative power (AUC=87.6%; Sensitivity=85.7% and Specificity=83.6% at the threshold of ≥1/4 points), with no evidence for over-optimism in bootstrapped internal validation.
Conclusion
These findings demonstrate the good diagnostic properties of a new screening scale using only 4 easy-to-collect items. If confirmed in other populations, it may prove useful in outpatient dermatology clinics for triage of psoriasis patients requiring further assessment by the rheumatologist.
This article is protected by copyright. All rights reserved.
