Background: Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) is a spectrum of high-mortality severe drug reactions. This study aimed to develop an in-hospital mortality risk calculator for SJS based on comorbidities for early risk prediction.
Methods: SJS/TEN patients were identified in National/Nationwide Inpatient Sample between Q4 2015-2020. Weight for each comorbidity was determined from a multivariable logistic regression to develop SJS Index.
Results: SJS Index had good discrimination power (c-statistic=0.704) and was well-calibrated (Brier score=0.049). SJS index had significantly better discriminative power than Elixhauser Comorbidity Index (p-value=0.001). SJS Index has good applicability in SJS-TEN and TEN. After adjusting for demographics, SJS Index had improved performance in all groups.
Conclusions: SJS Index effectively discriminates and predicts in-hospital mortality in SJS/TEN.
Development and Validation of a Comorbidity Index to Predict In-hospital Mortality Risk for Stevens-Johnson Syndrome
19th February 2026
PMID: 41742384
| Authors | Affiliations |
|---|---|
| Renxi Li | The George Washington University School of Medicine and Health Sciences, Washington, DC, USA Department of Surgery, University of Minnesota Medical School, Minneapolis, MN, USA |
Development and Validation of a Comorbidity Index to Predict In-hospital Mortality Risk for Stevens-Johnson Syndrome
Cite this article as:
Li R. Development and Validation of a Comorbidity Index to Predict In-hospital Mortality Risk for Stevens-Johnson Syndrome. Acute Med. 2025;24(2):86-92. doi: 10.52964/AMJA.1011. PMID: 41742384.
