Progression of drug exanthemas to serious drug eruptions: A retrospective review identifying early determinants.

BACKGROUND/OBJECTIVES: Maculo-papular drug exanthema (MPE) is the most common type of cutaneous adverse drug reaction (CAR). Exanthematous macules and papules may also be the initial presentation of severe CAR (SCAR). We aimed to identify characteristics associated with the diagnosis of SCAR in CAR-hospitalised patients. METHODS: This cross-sectional study was performed in a tertiary hospital in Chile. All CAR patients who were initially evaluated for exanthematous macules and papules were assessed for clinical, laboratory and pathological variables and these were contrasted with MPE or SCAR diagnosis at discharge. RESULTS: We enrolled 86 patients, of whom 25 (29%) had an at-discharge diagnosis of SCAR. SCAR patients were younger and the latency (time from starting drug to development of first skin lesions) was longer than in MPE patients: 43.6 +/- 18.7 years versus 54.0 +/- 21.8 years (P = 0.039) and 14 days; range 1 to 35, versus 7 days; range 1 to 45 (P = 0.001). The presence of cutaneous pain (OR 7.4 95% CI 1.3-41), mucosal involvement (OR 9.5 CI 95% 2.6- 34.5) and anticonvulsant use (OR 6.11 95% CI 1.91-19.53) were significantly associated with SCAR at discharge. Antibiotics use was significantly associated with MPE diagnosis (OR 2.8 95% CI 1.1-7.6). These six variables together explain 45% of the risk of having SCAR (R2 = 0.449). None of the early laboratory or pathological variables was associated with SCAR. CONCLUSIONS: In hospitalised patients assessed for exanthematous macules and papules, the evaluation of these clinical features may aid in the early identification of SCAR cases.