Stevens–Johnson Syndrome and Toxic Epidermal Necrolysis: A Five-Year Experience at a Tertiary Referral Hospital in Somalia

Abstract

Purpose: To describe the clinical characteristics, suspected drug triggers, management, and in-hospital outcomes of patients with Stevens-Johnson syndrome (SJS), SJS/toxic epidermal necrolysis (TEN) overlap, or TEN at a tertiary referral hospital in Somalia. Patients and Methods: This retrospective descriptive study included 37 patients admitted from January 2020 through December 2025. Age was categorized as 1-18, 19-49, and ≥50 years. Final clinician-documented diagnoses, demographic and clinical characteristics, suspected drugs, first-admission laboratory classifications, documented treatments, intensive care unit (ICU) admission, complications, length of stay, and survival were summarized using counts and percentages. Standardized SCORTEN and ALDEN assessments were unavailable. Mortality modeling was not performed because only seven deaths occurred and the extracted data were insufficient for a stable multivariable model. Results: Of 37 patients, five (13.5%) were aged 1-18 years, 18 (48.6%) were aged 19-49 years, and 14 (37.8%) were aged ≥50 years; 23 (62.2%) were female. Final clinician-documented diagnoses were TEN in 27 (73.0%), SJS/TEN overlap in six (16.2%), and SJS in four (10.8%). Antibacterial agents accounted for 26 (70.3%) suspected exposures, including other beta-lactams, trimethoprim-sulfamethoxazole, ceftriaxone, and amikacin. Laboratory findings indicated significant systemic involvement, including elevated inflammatory markers and liver enzymes. Prednisolone was widely used (94.6%), and more than half of patients received cyclosporine (56.8%), while IVIG use was minimal (2.7%). Nine patients (24.3%) required ICU admission. The overall in-hospital mortality rate was 18.9% (7/37). Conclusion: This single-center cohort was characterized by frequent suspected antibacterial drug exposure, a high proportion of clinician-documented TEN, and substantial in-hospital mortality. Because the study was small and retrospective and lacked standardized severity and drug-causality assessments, it cannot establish treatment effects or mortality predictors. Prospective multicenter studies using SCORTEN, ALDEN, standardized treatment documentation, and long-term follow-up are needed.