derm.sjs-ten.core.v1

Stevens-Johnson syndrome / toxic epidermal necrolysis (SJS/TEN)

dermatologyacuteadultpediatricacuteinpatient

Start encounter →Print handoutPRODUCTION

shard-06 Phase C gap-fill 2026-05-17 — first dermatology dossier in scope; SJS/TEN drug-eruption SCAR spectrum, ACUTE-CRITICAL. SCORTEN, day-3 SCORTEN, ABCD-10, and ALDEN have NO clinical-tools-registry calculator and are intentionally encoded in prose + severity_triggers + the co-located brief/bundle; the WIRED decision surface is workup.sjs_ten (resolves: clinical-tools-registry.ts adapter_id "sjs_ten") + calc.news2 (deterioration) + calc.qsofa / protocol.septic_shock (denuded-skin sepsis → id.sepsis.core.v1). Outpatient setting deliberately excluded — SJS/TEN is never managed outpatient; settings = acute + inpatient; setting_playbooks = ed → inpatient → icu(burn unit). RxNav reachable this session — all 5 regimen drug RxCUIs verified live 2026-05-17 against RxNav REST (/REST/rxcui/<cui>/property.json): cyclosporine 3008, etanercept 214555, IVIG 42386 (RxNorm "immunoglobulins, intravenous"), methylprednisolone 6902, prednisone 8640. Step 1 (culprit withdrawal + supportive bundle) is non_pharm. Culprit-class exemplars (allopurinol/carbamazepine/lamotrigine/sulfamethoxazole/nevirapine/oxicam NSAIDs) referenced in prose only — they are drugs to STOP/avoid, not regimen members. evidence.pmids first entry 27317286 is the Creamer 2016 UK BAD/BBA SJS/TEN guideline (Br J Dermatol 2016) — PubMed-verified 2026-05-22 (prior placeholder 27316286 resolved to an unrelated Can J Ophthalmol madarosis case report and was corrected); all 18 PMIDs are the established SCAR canon, every one PubMed-reverse-verified 2026-05-22, with effect sizes captured in _research-bundle.md Section 2. Cross-engine routing edges (by engine_id): sepsis-from-denuded-skin → id.sepsis.core.v1 (exists); ocular-surface salvage → ophthalmology consult / ophtho.ocular-surface-disease.core.v1 (not yet authored — sibling routing target); DRESS/AGEP/EM-major/SSSS/autoimmune-bullous siblings (not yet authored — encoded as sibling_differentiation with pivot findings + LRs). Bayesian linkage (pre-test priors per differential, 12 cardinal LR pairs incl Nikolsky/mucosal/biopsy-full-thickness-necrosis, 5 mimic-discrimination LRs, T_test/T_treat/T_disposition/T_immunomodulation thresholds, 5 cross-dossier routing edges) documented in _research-bundle.md §3; first-class TS fields remain schema-blocked (same pattern as id.cellulitis.core.v1). INTEGRATED (not PRODUCTION): terminology codes are plausible but not pipeline-validated; SCORTEN/ABCD-10 lack a wired calculator (prose-only); the Creamer guideline PMID needs source review. PRODUCTION blockers: terminology validation + Creamer PMID confirmation + dedicated SJS/TEN contract test.

Entry points (4)

symptom
Painful dusky macules / atypical targets → flaccid bullae + ≥2 mucosal sites (Bastuji-Garin classification, Arch Dermatol 1993; PMID 8420497)
painful_blistering_rash_with_mucositis
medication
New high-risk drug (allopurinol, aromatic anticonvulsant, sulfonamide, nevirapine, oxicam NSAID) started 4–28 d prior (Sassolas ALDEN, Clin Pharmacol Ther 2010; PMID 20375998; Mockenhaupt EuroSCAR, PMID 17805350)
new_high_risk_drug_4_to_28d
symptom
Prodromal fever / odynophagia / photophobia 1–3 d before rash + oral/ocular/genital erosions (Creamer UK guideline, Br J Dermatol 2016; PMID 27317286)
mucosal_erosions_with_prodrome
symptom
Positive Nikolsky sign + sheet-like epidermal detachment (Bastuji-Garin classification; PMID 8420497)
positive_nikolsky_sheet_detachment

Required inputs (18)

drug_exposure_historyrequired
medication • used at CONTEXT
ALDEN causality scoring — identify single most-probable culprit by latency 4–28 d (Sassolas ALDEN, Clin Pharmacol Ther 2010; PMID 20375998)
mucosal_involvementrequired
symptom • used at ENTRY
≥2 mucosal sites is a SJS/TEN-defining feature and differentiates from SSSS (mucosa spared) (Bastuji-Garin classification; PMID 8420497)
bsa_detachment_pctrequired
symptom • used at INITIAL_WORKUP
Epidermal-detachment %BSA defines spectrum tier (SJS <10 / overlap 10–30 / TEN >30) and is a SCORTEN variable (Bastuji-Garin; SCORTEN, J Invest Dermatol 2000; PMID 10951229)
ocular_involvementrequired
symptom • used at RED_FLAGS
Acute ocular-surface involvement mandates EMERGENT ophthalmology (amniotic-membrane window <7 d) (Creamer UK guideline; PMID 27317286)
temperaturerequired
vital • used at CONTEXT
Pyrexia from extensive skin loss vs sepsis; thermoregulation in denuded skin (Creamer UK guideline; PMID 27317286)
hrrequired
vital • used at RISK_STRATIFICATION
HR >120 is a SCORTEN variable; tachycardia signals sepsis/hypovolemia from skin loss (SCORTEN, J Invest Dermatol 2000; PMID 10951229)
sbprequired
vital • used at RED_FLAGS
Hypotension → distributive/hypovolemic shock from denuded skin → sepsis pathway (Creamer UK guideline; PMID 27317286)
hiv_status
history • used at CONTEXT
HIV raises SJS/TEN incidence ~100-fold (esp. sulfonamides, nevirapine) (Mittmann, Br J Dermatol 2012; PMID 22145749)
prior_scar
history • used at CONTEXT
Prior SCAR or culprit-class exposure → never re-expose; informs ALDEN (Sassolas ALDEN; PMID 20375998)
hla_risk_allele
history • used at CONTEXT
HLA-B*15:02 (carbamazepine/aromatic AEDs) and HLA-B*58:01 (allopurinol) are strong genetic risk markers (Chung Nature 2004 PMID 15057820; Hung PNAS 2005 PMID 15743917; CPIC PMID 29392710)
pregnancy
demographic • used at TREATMENT
Pregnancy alters immunomodulator choice (cyclosporine acceptable; avoid etanercept unless benefit outweighs) (Creamer UK guideline; PMID 27317286)
bunrequired
lab • used at INITIAL_WORKUP
BUN >10 mmol/L is a SCORTEN variable (SCORTEN, J Invest Dermatol 2000; PMID 10951229)
bicarbonaterequired
lab • used at INITIAL_WORKUP
Serum bicarbonate <20 mmol/L is a SCORTEN variable (SCORTEN; PMID 10951229)
glucoserequired
lab • used at INITIAL_WORKUP
Glucose >14 mmol/L is a SCORTEN variable (SCORTEN; PMID 10951229)
wbcrequired
lab • used at INITIAL_WORKUP
Baseline + serial CBC for infection surveillance (NO prophylactic antibiotics — culture-directed only) (Creamer UK guideline; PMID 27317286)
creatininerequired
lab • used at TREATMENT
Renal function gates cyclosporine dosing/contraindication and informs ABCD-10 (dialysis) (González-Herrada, J Invest Dermatol 2017; PMID 28634032; Noe ABCD-10, JAMA Dermatol 2019; PMID 30840032)
alt
lab • used at BRANCHING_WORKUP
Transaminitis flags DRESS overlap (eosinophilia + organ involvement) vs pure SJS/TEN (Mockenhaupt EuroSCAR; PMID 17805350)
biopsy_resultrequired
imaging • used at BRANCHING_WORKUP
Lesional skin biopsy — full-thickness epidermal necrosis + DIF-negative confirms SJS/TEN and excludes SSSS (intraepidermal) and autoimmune bullous (DIF+) (Bastuji-Garin; PMID 8420497; Creamer; PMID 27317286)

12-phase flow (12)

1FRAME
Position on the SJS↔TEN continuum by %BSA epidermal detachment (SJS <10%, overlap 10–30%, TEN >30%); SJS/TEN is one disease, EM-major/SSSS/autoimmune-bullous are distinct (Bastuji-Garin consensus classification, Arch Dermatol 1993; PMID 8420497)
advance: spectrum position framed; scope is drug-induced SCAR continuum
2ENTRY
Recognize painful skin + ≥2 mucosae + new high-risk drug 4–28 d prior (Sassolas ALDEN, Clin Pharmacol Ther 2010; PMID 20375998; Mockenhaupt EuroSCAR; PMID 17805350)
inputs: mucosal_involvement
advance: severe cutaneous adverse reaction suspected
3CONTEXT
ALDEN causality per suspect drug (latency 4–28 d "probable"); HIV / prior SCAR / HLA-B*15:02 / HLA-B*58:01 risk; comorbidity (Sassolas ALDEN; PMID 20375998; Mittmann; PMID 22145749; CPIC; PMID 29392710)
inputs: drug_exposure_history, hiv_status, prior_scar, hla_risk_allele, temperature
advance: single most-probable culprit identified via ALDEN
4RED_FLAGS
Nikolsky+/Asboe-Hansen+, sheet detachment, ocular/airway/genitourinary mucosal sloughing, rapid progression, sepsis physiology, SCORTEN-high — STAT critical-care + multidisciplinary consults (Bastuji-Garin; PMID 8420497; SCORTEN; PMID 10951229; Creamer; PMID 27317286)
inputs: bsa_detachment_pct, mucosal_involvement, ocular_involvement, sbp
actions: calc.news2, calc.qsofa, protocol.septic_shock
advance: critical features triaged and consults initiated
5INITIAL_WORKUP
%BSA detachment mapping; lesional + perilesional skin biopsy (frozen H&E ± DIF); CBC / renal / LFT / glucose / bicarbonate; blood + skin cultures; SCORTEN 7-variable labs (Creamer UK guideline; PMID 27317286; SCORTEN; PMID 10951229)
inputs: bsa_detachment_pct, bun, bicarbonate, glucose, wbc
actions: workup.sjs_ten, panel.cbc, panel.renal, panel.lft
advance: biopsy sent + SCORTEN variables resulted
6BRANCHING_WORKUP
Biopsy resolves the bullous differential: full-thickness epidermal necrosis + subepidermal split + DIF-negative = SJS/TEN; intraepidermal granular-layer split = SSSS; DIF-positive = autoimmune bullous; eosinophilia + LFT/Cr derangement + facial edema = DRESS overlap (Bastuji-Garin; PMID 8420497; Mockenhaupt EuroSCAR; PMID 17805350)
inputs: biopsy_result, alt
actions: workup.sjs_ten
advance: histologic pattern + overlap labs confirm terminal dx
7DIFFERENTIAL
SJS/TEN vs DRESS vs AGEP vs erythema multiforme major vs SSSS vs autoimmune bullous (pemphigus/bullous pemphigoid/paraneoplastic) vs acute GVHD vs generalized bullous fixed drug eruption vs MIRM (pediatric) (Mockenhaupt EuroSCAR; PMID 17805350; Canavan MIRM, J Am Acad Dermatol 2015; PMID 25592340)
inputs: biopsy_result, mucosal_involvement
advance: terminal diagnosis assigned
8RISK_STRATIFICATION
SCORTEN (7-variable, day-1 AND day-3 recompute) + ABCD-10 → mortality estimate + disposition tier (Bastuji-Garin SCORTEN, J Invest Dermatol 2000; PMID 10951229; Guégan day-3 SCORTEN; PMID 16374461; Noe ABCD-10, JAMA Dermatol 2019; PMID 30840032)
inputs: bsa_detachment_pct, bun, bicarbonate, glucose, hr
actions: calc.news2
advance: SCORTEN + ABCD-10 documented with predicted mortality
9TREATMENT
Day-0 IMMEDIATE culprit withdrawal (the single highest-impact intervention — delay raises mortality; Garcia-Doval, Arch Dermatol 2000; PMID 10724193) + supportive critical care (fluid/electrolyte, non-adherent wound care, NO prophylactic antibiotics, temperature/airway/ocular protection, analgesia, nutrition; Creamer UK guideline; PMID 27317286) ± CONTESTED systemic immunomodulation: cyclosporine (SMR 0.42; González-Herrada; PMID 28634032) or etanercept (SMR 0.51 vs steroid 0.91; Wang RCT, J Clin Invest 2018; PMID 29400697) or IVIG ± short-pulse steroid (no robust survival benefit; Barron; PMID 24697283; Zimmermann meta-analysis; PMID 28329382)
inputs: bsa_detachment_pct, creatinine, pregnancy
advance: culprit stopped + supportive plan + immunomodulation decision documented
10DISPOSITION
Burn unit / ICU if detachment >10% BSA OR rapid progression OR SCORTEN ≥2 OR ABCD-10 ≥2; early transfer (<24 h) lowers mortality; place ophthalmology / urology-gynecology / airway consults (Palmieri, J Burn Care Rehabil 2002; PMID 11882797; Creamer UK guideline; PMID 27317286)
inputs: bsa_detachment_pct
advance: disposition assigned + multidisciplinary consults placed
11MONITORING
Daily %BSA + re-epithelialization, fluid/electrolyte/glucose, infection surveillance (culture-directed, NOT prophylactic), day-3 SCORTEN recompute, ophthalmology review q24h while acute (Creamer UK guideline; PMID 27317286; Guégan; PMID 16374461)
inputs: bsa_detachment_pct, wbc, creatinine
actions: panel.cbc, panel.renal
advance: re-epithelialization underway and no new mucosal/organ progression
12FOLLOWUP
Ocular (symblepharon, dry eye, vision loss), cutaneous (dyspigmentation, scarring), oral, genital (synechiae), pulmonary (bronchiolitis obliterans) sequelae screen; psychological support; HLA pharmacogenomic counseling + lifelong culprit-class + cross-reactor avoidance card (Creamer UK guideline; PMID 27317286; CPIC HLA-B guideline, Clin Pharmacol Ther 2018; PMID 29392710)
advance: sequelae plan + pharmacogenomic avoidance card finalized