Validation of PARACELSUS score performance for the diagnosis of pyoderma gangrenosum: An international multicenter study with 1403 cases.

Moelleken, Maurice; Ortega-Loayza, Alex G; Busch, Dorothee; Erfurt-Berge, Cornelia; Shinohara, Michi M; Do, An; Libertad Ortega Nuñez, Diana; Vega, Mineli Díaz et al. · J Am Acad Dermatol · 2026

retrospective_cohort · Level III

Where this comes from

Abstract

Pyoderma gangrenosum is a rare ulcerative neutrophilic dermatosis which is challenging to diagnose due to heterogeneous presentation and overlap with other ulcerative diseases. Currently, PARACELSUS score is the most sensitive diagnostic tool, but its specificity is limited. Evaluation of PARACELSUS score for diagnosis of diverse wound etiologies. International, multicenter, retrospective observational diagnostic accuracy validation study included 1403 cases from 14 institutions in 7 countries (2022-2024). All cases had a confirmed diagnosis of pyoderma gangrenosum or other wound entities. Diagnostic accuracy statistics were applied to PARACELSUS score using established and potential cut-off values. Among 1403 cases (57.0% women, 43.0% men; mean age 62 years), 85 wound entities were identified, including 180 cases of PG. Raising the diagnostic cut-off from ≥10 to >10 points improved specificity (93.2% vs 96.8%; P < .001), positive predictive value (68.4% vs 81.9%; P < .001) and accuracy (94.1% vs 97.0%; P < .001). The false-positive rate was decreased (6.8% vs 3.2%; P < .001) with a non-significant reduction in sensitivity (100.0% vs 98.3%; P = .25). Retrospective data collection and heterogeneity of participating centers may limit generalizability. Adjusting PARACELSUS score cut-off to >10 points significantly improves specificity and accuracy while maintaining high sensitivity, thereby reducing false-positive diagnoses.