Standardising Assessment of Superficial Pyoderma in Dogs: Proposal of a Severity Score

Background Superficial pyoderma is common in dogs with atopic dermatitis, yet no validated and widely accepted severity scoring system exists. Hypothesis/Objectives To develop a predictor for superficial pyoderma severity based on reliable objective lesion quantification combined with a clinician‐assigned numerical severity rating. Animals Sixty client‐owned dogs presenting with follicular papules, pustules, epidermal collarettes, and/or yellow crusts were prospectively enrolled. Materials and Methods Six clinicians assessed lesion types, numerical and categorical disease severity, and treatment efficacy at baseline and at up to two follow‐up visits. Lesions were quantified by counting individual papules and pustules, and by standardised area measurements for confluent papules/pustules, epidermal collarettes and crusts. Inter‐ and intra‐clinician reliability were evaluated. A predictive model for pyoderma severity was subsequently developed using lesion quantification and clinician‐assigned numerical severity scores. Results Recognition of specific lesion types and categorical severity assessments showed poor agreement between clinicians, indicating limited reliability for severity assessment. By contrast, combined lesion quantification demonstrated good inter‐ and intra‐clinician reliability, and clinician‐assigned numerical severity scores showed good inter‐clinician reliability. The resulting predictive model demonstrated good agreement with clinician‐assigned severity scores, estimating pyoderma severity on a continuous scale from 0 (no pyoderma) to 100 (worst conceivable pyoderma), with a mean absolute error of approximately nine severity points per case. Conclusions and Clinical Relevance The proposed scoring system relies on combined lesion quantification without the need for subjective categorical grading. Although further validation is required, it provides a foundation for objective assessment of superficial pyoderma severity and introduces a mathematical model that translates combined lesion burden into clinician‐perceived disease severity.