Background: There are no standardized measures for the severity of rosacea.
Objectives: To determine what clinical signs correlate best with global severity of rosacea, and to examine inter-rater reliability. Methods Four clinicians each made 82 assessments of rosacea patients. Each assessment used 60 0-to-10 Likert-like scales for the signs of rosacea. Subjects also assessed the severity of their rosacea.
Results: The clinicians' assessment of global severity correlated strongly with erythema, especially on the cheeks. Subjects' assessment of global severity correlated more strongly with papules/pustules. Different methods of assessing severity--estimation of area involved, intensity, or lesion counts--did not produce significantly different results. Inter-rater reliability was low on 11-point (0-10) scales, but improved when scales were collapsed to 5 or 4 points.
Conclusions: Clinicians and patients assess severity of rosacea differently, with clinicians focusing on erythema and patients focusing on papules/pustules. New instruments for assessing severity must address inter-rater reliability.