Clinical and Demographic Predictors of In‐Hospital Mortality in Adult SJS/TEN Hospitalizations: A National Inpatient Sample Analysis 2017–2022

ABSTRACT Background Stevens−Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe mucocutaneous reactions associated with significant morbidity and mortality. Published studies have identified a limited number of variables predictive of death, but their description of acute complications is sparse. Objectives To analyse adult hospitalizations for SJS/TEN using a national claims‐based database, characterize associated acute complications and determine how these complications influence in‐hospital mortality. Methods We retrospectively identified adult hospitalizations with SJS/TEN from the 2017−2022 National Inpatient Sample (NIS) Database. Variables with p ≤ 0.2 in univariable screening were included in the multivariable logistic regression, with significance set at p ≤ 0.05. Results Among 29,875 adult SJS/TEN hospitalizations over the 6‐year study period, 7.9% died in the hospital. On univariable analysis, the strongest predictors of mortality were disseminated intravascular coagulation (DIC) (OR 19.15), respiratory failure (OR 15.79), anoxic brain injury (OR 12.38) and hepatic failure (OR 9.86); all p < 0.001. African American race, Medicare insurance, aspiration pneumonia, systolic heart failure, myocardial infarction, hyponatremia and pulmonary embolism lost significance after multivariable adjustment. On multivariable analysis, the following factors were associated with in‐hospital death: age per year (OR 1.03), Charlson Comorbidity Index (CCI) per point (OR 1.11; 95% CI 1.062–1.168), acidosis (OR 2.49), acute kidney injury (AKI) (OR 1.96; 95% CI 1.509–2.557), anoxic brain damage (OR 4.77), DIC (OR 4.16; 95% CI 2.439–7.089), encephalopathy (OR 1.38; p = 0.032), hepatic failure (OR 3.18; 95% CI 1.999–5.066), infection (OR 1.95) and respiratory failure (OR 6.80; 95% CI 5.250–8.805); all p < 0.001 unless otherwise noted. Limitations This retrospective analysis relies on an administrative database with diagnoses identified by billing codes, which are subject to misclassification and it lacks data on disease severity (body surface area involvement), culprit drugs, laboratory values and treatment, limiting adjustment for confounding by indication. Only in‐hospital mortality was assessed, not longer‐term outcomes. Conclusions In‐hospital mortality among adult SJS/TEN patients was 7.9% and was associated with older age, greater comorbidity burden and the presence of clinical complications including acidosis, AKI, anoxic brain damage, DIC, encephalopathy, hepatic failure, infection and respiratory failure. Early identification and proactive management of high‐risk patients, particularly those with organ dysfunction, may be critical to reducing mortality in this vulnerable population. These results offer actionable insights for clinicians to reduce the mortality of SJS/TEN patients.