Clinical Commander

All dossiers
neuro.ischaemic-stroke.v1

Acute Ischaemic Stroke

neurologyacuteadultacuteinpatient
Start encounter →Print handoutPRODUCTION

Phase B deepening (2026-05-14): TOAST phenotypes (LAA/CE/SVO/ODE/UDE) encoded as severity_triggers — pivot from SiblingDifferentiation rows because future LAA/CE/SVO/ODE/UDE engines are Phase C expansion targets and sibling_engine_id requires registry resolution. 5 setting playbooks span the full journey: home (prehospital BE-FAST + LAMS/RACE + transport-destination) → ed (door-to-needle) → icu (post-reperfusion) → inpatient (stroke unit + secondary prevention init) → outpatient (90-d / 12-mo follow-up). Schema-blocked calculators surfaced as tickets in docs/framework-audit/shard-3-neuro-sym-state.md: calc.aspects (Alberta Stroke Program Early CT Score 0–10, ≥6 EVT-eligible), calc.mrs (modified Rankin 0–6, baseline + 90-d + 12-mo), calc.toast (5-class classification system), calc.guss (Gugging swallow screen). Currently encoded as plain-English required_assessments and severity_triggers until registry entries land. TIA-engine (calc.abcd2 host) not yet built — calculators[calc.abcd2] flagged `required: false` until cardio.tia.v1 promotes from PLANNED. Regimen axes encoded: 5-step acute reperfusion (TNK/alteplase → EVT → BP control labetalol/nicardipine/clevidipine → 24-h antiplatelet → secondary prevention atorvastatin/DOAC). All RxCUIs RxNav-validated (none touched this pass). EVIDENCE CORRECTION (2026-05-18): the prior depth.md "12-PMID anchor set" was UNVERIFIED — PubMed-MCP verification found 8/12 mis-attributed to unrelated articles (35779546=NHS privatisation→AcT 35779553; 36369796=seawater chemistry→ATTENTION 36239644; 36369797=beagle-dog QTc→BAOCHE 36239645; 24449944=GBD 2010→DESTINY II 24645942; 25882509=Furlan editorial→EXTEND-IA 25671797; plus MR CLEAN/ESCAPE/SWIFT-PRIME/REVASCAT label swaps — true MR CLEAN 25517348, ESCAPE 25671798, EXTEND-IA 25671797, SWIFT-PRIME 25882376, REVASCAT 25882510). evidence.pmids replaced with the verified set; full correction table in src/lib/dossiers/neuro.ischaemic-stroke.v1._research-bundle.md §2. Only 7477192/18815396/26898852/24963567/29129157/29364767/23803136/29766750 were already correct. §5.5.1 effect sizes (each tied to a PubMed-MCP-verified PMID): NINDS (PMID 7477192) alteplase ≤3 h ≥30% more likely minimal/no disability, sICH 6.4% vs 0.6%; ECASS-III (18815396) 3-4.5 h mRS 0-1 52.4% vs 45.2% (OR 1.34, 1.02-1.76); AcT (35779553) TNK non-inferior, mRS 0-1 36.9% vs 34.8% (RD 2.1%, 95% CI -2.6 to 6.9), n=1600, sICH 3.4% vs 3.2%; HERMES (26898852) anterior-LVO EVT aOR 2.49, NNT 2.6; ESCAPE (25671798) functional independence 53.0% vs 29.3%, mortality 10.4% vs 19.0%; EXTEND-IA (25671797) 71% vs 40%; SWIFT-PRIME (25882376) 60% vs 35%; DAWN (29129157) 6-24 h EVT 49% vs 13%; DEFUSE-3 (29364767) 6-16 h EVT 45% vs 17%, mortality 14% vs 26%; DESTINY II (24645942) hemicraniectomy age ≥61 survival-without-severe-disability 38% vs 18% (OR 2.91), mortality 33% vs 70%; CRYSTAL-AF (24963567) ILR AF detection 12.4% vs 2.0% at 12 mo (HR 7.3). §5.5.2 ros-and-ddx layer added (auto-registered by readdir): neuro.ischaemic-stroke.v1.ros.ts (13 ROS), .differentials.ts (14 TOAST-stratified AIS + must-not-miss mimics with cohort-sourced pre-test priors), .finding-lrs.ts (23 LR rows, REWRITE of the prior thin 98-line file — 3 prior rows kept & now PMID-verified). Named pivots wired: stroke vs ICH (CT hyperdensity), vs hypoglycaemia (fingerstick + glucose response), vs seizure-Todd (positive-vs-negative phenomena + march), vs Bell (forehead pivot), vs migraine-aura (spread tempo). 2 conditional-dependency notes: NIHSS as a correlated composite band (not component-multiplied; Demeestere 2017 PMID 28457621); CTP/clinical-core mismatch LR conditional on time-since-LKW window (DAWN 6-24 h / DEFUSE-3 6-16 h). Resolving cross-dossier routes (engine_ids verified on disk 2026-05-18): neuro.ich.core.v1 (CT-blood pivot — never thrombolyse), neuro.sah.core.v1 (thunderclap), neuro.tia.v1 (tissue-negative, ABCD2 + DAPT), neuro.stroke-laa.v1 / neuro.stroke-cardioembolic.v1 / neuro.stroke-svo.v1 (TOAST mechanism-specific secondary prevention), neuro.first-seizure-eval.v1 (Todd), neuro.bell-palsy.v1 (forehead-involving isolated VII), neuro.migraine.core.v1 (migrainous aura). Special populations (2026 AHA/ASA, as data): (1) Pregnancy — pregnancy is NOT an absolute thrombolysis contraindication; IV thrombolysis reasonable when benefit outweighs uterine-bleeding risk; EVT preferred for LVO; multidisciplinary OB/neuro decision. (2) Pediatric AIS — the 2026 AHA/ASA guideline carries the FIRST-EVER pediatric acute-stroke recommendations; pediatric stroke pathway is a distinct workup (not adult dosing); flagged for a future neuro.peds-stroke engine. (3) Elderly — age alone does not exclude thrombolysis or EVT (HERMES showed benefit ≥80 y, cOR 3.68); hemicraniectomy benefit in age ≥61 is survival-with-disability (DESTINY II 24645942) — explicit goals-of-care discussion. (4) Wake-up / extended window — 2026 AHA/ASA endorses thrombolysis 4.5-9 h and 4.5-24 h from LKW in SELECT patients with perfusion mismatch ratio ≥1.2 or ≥1.8 (extends the prior ≤4.5 h tPA / DAWN-DEFUSE-3 EVT windows); MRI DWI-FLAIR mismatch for unwitnessed/wake-up onset. (5) BP/glucose pre-tPA — SBP must be <185/110 pre-thrombolysis and <180/105 ×24 h post (unchanged); correct hypoglycaemia <60 mg/dL FIRST and avoid hyperglycaemia (target 140-180 mg/dL).

Entry points (4)

  • symptom
    Sudden focal neurological deficit (FAST/BEFAST; AHA/ASA 2019)
    sudden_focal_deficit
  • symptom
    Aphasia / dysarthria of sudden onset (AHA/ASA 2019)
    aphasia
  • symptom
    Hemiparesis / facial droop (AHA/ASA 2019)
    hemiparesis
  • imaging
    Large-vessel occlusion on CTA (AHA/ASA 2019 Class I)
    lvo_on_cta

Required inputs (14)

  • agerequired
    demographic • used at CONTEXT
    Age affects EVT eligibility and bleeding risk (AHA/ASA 2019; DESTINY II, Jüttler NEJM 2014 — age <60 for hemicraniectomy)
  • last_known_wellrequired
    symptom • used at TREATMENT
    LKW determines tPA/TNK window ≤4.5 h (ECASS III, Hacke NEJM 2008) and EVT window ≤6 h or 6–24 h with mismatch (DAWN, Nogueira NEJM 2018; DEFUSE 3, Albers NEJM 2018)
  • nihssrequired
    symptom • used at RISK_STRATIFICATION
    NIHSS gates thrombolysis and EVT eligibility (≥6 for EVT; AHA/ASA 2019 Class I)
  • sbprequired
    vital • used at TREATMENT
    SBP <185/110 pre-tPA; <180/105 × 24 h post-tPA (AHA/ASA 2019 Class I); <220/120 if no thrombolysis
  • glucoserequired
    vital • used at CONTEXT
    Hypoglycaemia mimics stroke; hyperglycaemia worsens outcome (AHA/ASA 2019 Class I)
  • inrrequired
    lab • used at TREATMENT
    INR >1.7 contraindicates thrombolysis per AHA/ASA 2019 exclusion criteria
  • platelet_countrequired
    lab • used at TREATMENT
    Platelets <100k contraindicates thrombolysis per AHA/ASA 2019 exclusion criteria
  • creatininerequired
    lab • used at CONTEXT
    eGFR for contrast (CTA/CTP) and DOAC dose adjustment (AHA/ASA 2019)
  • ct_head_noncontrastrequired
    imaging • used at INITIAL_WORKUP
    Rule out hemorrhage before thrombolysis; ASPECTS scoring (AHA/ASA 2019 Class I)
  • cta_head_neckrequired
    imaging • used at INITIAL_WORKUP
    LVO identification for EVT triage (AHA/ASA 2019 Class I)
  • ct_perfusion_or_dwi_pwi
    imaging • used at BRANCHING_WORKUP
    Mismatch selection for 6–24 h window (DAWN, Nogueira NEJM 2018; DEFUSE 3, Albers NEJM 2018)
  • current_anticoagulantrequired
    medication • used at TREATMENT
    DOAC last dose; need anti-Xa or DTI levels; affects tPA eligibility (AHA/ASA 2019)
  • atrial_fibrillation
    history • used at FOLLOWUP
    AF as cardioembolic source → anticoagulation timing decision (2024 ESC AF guidelines)
  • recent_surgery_or_bleedrequired
    history • used at TREATMENT
    Major surgery <14 d / GI/GU bleed <21 d are tPA contraindications (AHA/ASA 2019)

12-phase flow (12)

  1. 1FRAME
    Acute focal deficit suggesting ischaemia — exclude hemorrhage, mimic, post-ictal (AHA/ASA 2019)
    advance: stroke alert activated
  2. 2ENTRY
    Capture LKW + activate door-to-needle timer (AHA/ASA 2019 Class I — door-to-needle ≤60 min)
    inputs: age, last_known_well
    advance: entry trigger captured
  3. 3CONTEXT
    Vitals, glucose, anticoagulants, recent surgery/bleeding, baseline mRS (AHA/ASA 2019)
    inputs: sbp, glucose, creatinine, current_anticoagulant, recent_surgery_or_bleed
    advance: context captured + tPA contraindication checklist run
  4. 4RED_FLAGS
    Hemorrhage on CT, malignant infarct/herniation signs, posterior circulation with rapid decline, BP > thrombolysis ceiling (AHA/ASA 2019)
    inputs: sbp
    advance: no immediate exclusion
  5. 5INITIAL_WORKUP
    Non-contrast CT + CTA head/neck; CBC, CMP, coags; ECG; troponin; finger-stick glucose (AHA/ASA 2019 Class I)
    inputs: ct_head_noncontrast, cta_head_neck, inr, platelet_count
    actions: panel.renal
    advance: imaging back + thrombolysis contraindications cleared
  6. 6BRANCHING_WORKUP
    CT perfusion or MRI DWI/PWI for late-window selection (DAWN, Nogueira NEJM 2018; DEFUSE 3, Albers NEJM 2018); carotid US/MRA; TTE for cardioembolic source; long-term ECG monitoring (STROKE-AF)
    inputs: ct_perfusion_or_dwi_pwi
    advance: mismatch profile assessed
  7. 7DIFFERENTIAL
    TOAST classification (Adams Stroke 1993): large-artery atherothrombotic / cardioembolic / lacunar / cryptogenic / other
    advance: mechanism assigned
  8. 8RISK_STRATIFICATION
    NIHSS for severity + EVT eligibility; ASPECTS for parenchymal damage; mRS baseline (AHA/ASA 2019 Class I)
    inputs: nihss
    advance: NIHSS + ASPECTS documented
  9. 9TREATMENT
    IV thrombolysis (TNK 0.25 mg/kg ≤4.5 h preferred per AcT Lancet 2022, alteplase per NINDS 1995/ECASS III Hacke NEJM 2008); EVT for LVO ≤6 h Class I, 6–24 h with mismatch (DAWN/DEFUSE-3); BP control with labetalol/nicardipine (AHA/ASA 2019); DAPT for minor stroke / high-risk TIA (POINT/CHANCE)
    inputs: sbp, last_known_well, inr, platelet_count
    advance: reperfusion strategy executed or excluded; BP at target
  10. 10DISPOSITION
    Stroke unit / NCCU; transfer to comprehensive stroke centre if EVT needed and unavailable (AHA/ASA 2019 Class I)
    inputs: nihss
    advance: level-of-care set
  11. 11MONITORING
    Neuro checks q15 min × 2 h → q30 min × 6 h → q1 h × 16 h; BP per AHA/ASA 2019 protocol; continuous telemetry for AF
    inputs: sbp, nihss
    advance: protocol active
  12. 12FOLLOWUP
    Secondary prevention: high-intensity statin LDL <55 (2026 ACC/AHA Lipid), BP <130/80 (2025 AHA/ACC HTN), DAPT/anticoag per source, dysphagia/SLP, rehab, AF screen
    inputs: atrial_fibrillation
    advance: secondary-prevention bundle prescribed + outpatient stroke clinic scheduled