SCORTEN: Validated prognostic score for SJS and Toxic Epidermal Necrolysis.
Prognostic Indicators
Risk Prediction
Select the clinical and laboratory criteria present on Day 1 to calculate the survival probability using the SCORTEN model.
Guidelines & Evidence
Verified
Last Review: 2026-07-17
When to Use
When to Use
SCORTEN (SCORe of Toxic Epidermal Necrosis) is a validated 7-variable severity-of-illness score that predicts in-hospital mortality for patients admitted with Stevens-Johnson Syndrome (SJS) or Toxic Epidermal Necrolysis (TEN). It is calculated within the first 24 hours of admission and should be repeated on Day 3 for prognostic refinement. It is the globally accepted standard mortality risk tool for SJS/TEN.
Primary Indications
SJS: < 10% BSA epidermal detachment with mucosal involvement and target lesions
SJS-TEN overlap: 10–30% BSA epidermal detachment
TEN: > 30% BSA epidermal detachment (Nikolsky positive)
Triaging to burn unit, medical ICU, or specialized dermatology centre based on predicted mortality
Prognostic communication with patient, family, and multidisciplinary team
Benchmarking outcomes in SJS/TEN case series and quality improvement
When NOT to Use
DRESS syndrome — use RegiSCAR scoring; SCORTEN variables are not validated for DRESS
AGEP — distinct pathology; self-limited; SCORTEN is inappropriate
Fixed drug eruption or morbilliform drug rash without epidermal detachment
As a substitute for clinical monitoring — SCORTEN is prognostic, not a management protocol
How it Works
Risk Factors (1 Point Each)
| Age > 40 years | 1 point |
| Heart rate > 120 beats/min | 1 point |
| Underlying malignancy (active or recent) | 1 point |
| BSA detachment > 10% on Day 1 | 1 point |
| Serum urea > 10 mmol/L (> 28 mg/dL) | 1 point |
| Serum glucose > 14 mmol/L (> 252 mg/dL) | 1 point |
| Serum bicarbonate < 20 mmol/L (< 20 mEq/L) | 1 point |
Predicted In-Hospital Mortality
Pathophysiological Basis
SJS/TEN results from drug-specific cytotoxic T-lymphocyte (CTL) and NK-cell-mediated keratinocyte apoptosis, driven by granulysin secretion and FasL–Fas interactions. The resulting massive epidermal detachment produces a burn-equivalent wound with fluid and protein loss, thermoregulatory failure, and mucosal barrier breakdown. The SCORTEN variables capture systemic physiological decompensation: tachycardia (hypovolemia/systemic inflammation), hyperglycemia (stress response), metabolic acidosis (bicarbonate < 20), and azotemia (dehydration and renal compromise). These metabolic markers — rather than skin area alone — are the strongest independent predictors of mortality.
Clinical Pearls
Key Pearls
Tachycardia (HR > 120) is often the most rapidly deteriorating variable in the first 24 hours and should prompt immediate aggressive fluid resuscitation.
The BSA > 10% criterion on Day 1 frequently underestimates final detachment area — reassess at 48 and 72 hours and update SCORTEN accordingly.
Malignancy as a variable reflects both immunosuppression (impaired clearance of drug-reactive T cells) and the frequent use of high-risk drugs (allopurinol for tumour lysis, antiepileptics for seizure prophylaxis, sulfonamides for PCP prophylaxis).
Limitations
SCORTEN was validated only at specialized centres — predictions may not apply to general medical wards.
Glucose > 14 mmol/L is the least discriminating variable; pre-existing diabetes may inflate the score without reflecting the severity of SJS/TEN itself.
SCORTEN does not account for treatment interventions (ciclosporin, IVIG, etanercept) — scores reflect untreated natural history from the 1990s era.
The original derivation included both SJS (< 10% BSA) and TEN (> 30% BSA) — performance may differ across the spectrum.
Treatment Controversy
No systemic therapy for TEN has RCT-level evidence. Current practice involves ciclosporin (3–5 mg/kg/day), etanercept, or IVIG based on retrospective data and case series. Corticosteroids are generally avoided or used cautiously due to infection risk. Immediate drug withdrawal is the only universally recommended intervention.
Next Steps
Triage and Management by SCORTEN
| SCORTEN 0–1 (~3% mortality) | Admit to dermatology ward or step-down unit. Wound care. Discontinue all suspect drugs. Ophthalmology and urology review. Serial bloods q24h. |
| SCORTEN 2 (~12% mortality) | Consider burn unit or medical HDU. IV fluid resuscitation. Nutritional support (NG feeding if mucosal involvement limits oral intake). Ophthalmology — topical cyclosporine drops to prevent symblepharon. |
| SCORTEN 3–4 (35–58% mortality) | Burn unit or ICU mandatory. Aggressive fluid resuscitation (similar principles to burn care). Consider ciclosporin or etanercept (specialist decision). Intensive wound care — non-adherent dressings, avoid mechanical debridement. ICU monitoring. |
| SCORTEN ≥ 5 (> 90% mortality) | Burn ICU. Early goals-of-care conversation with patient/family. All above measures. Palliative care team involvement if patient/family elected comfort-focused care. |
Mandatory Immediate Actions (All Cases)
01
Stop ALL suspected drugs immediately — compile full medication history; allopurinol, antiepileptics, antibiotics, NSAIDs, nevirapine are commonest culprits
02
Establish IV access — fluid resuscitation with LR or Hartmann's solution to maintain urine output 0.5–1 mL/kg/hr
03
Ophthalmology consultation within 24 hours — ocular involvement occurs in 50–80% of TEN and can cause permanent blindness without early intervention
04
Urology review if urethral/genital mucosal involvement — catheterize early to prevent urethral stricture
05
Nutritional assessment — parenteral or NG feeding if oral intake impaired
06
Initiate ALDEN score or clinical causality assessment to identify and document the probable causative drug
07
Issue drug allergy card on discharge — patient must never receive the culprit drug or cross-reactive agents again
The Evidence
Original Derivation & Validation
SCORTEN: a severity-of-illness score for toxic epidermal necrolysis.
Bastuji-Garin S et al. • J Invest Dermatol. 2000;115(2):149-153. Derivation (n=165) and internal validation (n=75) cohort from Hôpital Henri Mondor burn unit, Paris. Seven independent prognostic variables identified; C-statistic 0.85 in derivation cohort.
View SourceExternal Links
Origins & History
Origins
SCORTEN was developed by Sylvie Bastuji-Garin and colleagues at Hôpital Henri Mondor, Créteil, France — a centre with one of the largest SJS/TEN patient volumes in Europe — and published in the Journal of Investigative Dermatology in 2000. The score emerged from recognition that existing severity measures (SAPS II, burn scoring systems) performed poorly in TEN, which shares features with both medical critical illness and burn injury but has a distinct biology and trajectory. Bastuji-Garin's group used logistic regression on 165 admitted patients to identify the seven independently prognostic variables and validate the mortality probability table. Despite its age, SCORTEN remains the most widely externally validated prognostic tool for SJS/TEN, having been validated in cohorts from France, Germany, Taiwan, and across multiple systematic reviews.
Last Comprehensive Review: 2026-07-17