RESUMO
OBJECTIVES: Skin reactions to drugs affect about 2.2 percent of inpatients. More often, they are described as morbilliform or maculopapular. The objectives of this study were to characterize these skin reactions and to look for severity markers. METHODS: We retrospectively analyzed 133 cases of drug reactions collected in a Dermatology unit from 1973 to 1995 for whom, photographs and blood cell count were available. Well recognized disorders were excluded (Stevens-Johnson syndrome, Lyell syndrome, TEN, AGEP, vasculitis, phototoxicity, fixed drug eruption). Severity was defined on 3 criteria: hospitalization, skin reaction prolonged more than 21 days and visceral involvement. For each criteria, we studied the association with clinical and laboratory findings, first with univariate analysis, then with multivariate analysis for significant associations. RESULTS: Characteristics of the 133 included cases were: women: 58 p. 100, mean age: 52 years, inpatients: 80 p. 100, mean interval from beginning of drug therapy to reaction: 12 days, mean duration: 26 days, responsible drugs: antibiotics: 41 p. 100, nonsteroidal antiinflammatory drugs: 11 p. 100, anticonvulsants: 10 p. 100. A few reactions showed a monomorphous pattern fitting a single denomination in a classic dermatological nomenclature: maculopapular: 10 p. 100, morbilliform: 6 p. 100, scarlatiniform: 5 p. 100, small papules: 5 p. 100. Polymorphous reactions were observed in 73 p. 100 of cases with 3 different patterns on the average. Nearly half of the cases exhibited maculopapular and scarlatiniform lesions simultaneously. We observed fever (61 p. 100), mucosal lesions (26 p. 100), lymphadenopathy (30 p. 100), eosinophilia (count > 500/mm3: 44 p. 100, > 1500/mm3: 17 p. 100), and visceral involvement (22 p. 100), including hepatitis (18 p. 100). Four severity markers were identified: fever, lymphadenopathy, erythema involving more than 60 p. 100 of the body surface area, eosinophilia. CONCLUSION: This study underlines the polymorphous clinical presentation of skin reactions to drugs and the minority of maculopapular or morbilliform patterns leading us to propose the term of "erythematous drug reactions". Despite several biases (hospital recruitment of more severe reactions, retrospective analysis) this study is the first to show some severity markers. Close to the criteria used to define the "hypersensitivity syndrome" these simple clinical markers may have important practical implications.