Clinical Commander

All dossiers
psych.delirium.core.v1

Delirium — acute confusional state (DSM-5-TR 2022; SCCM PADIS 2018 PMID 30113379; NICE CG103 2010 reaffirmed 2024; CAM Inouye 2006 PMID 16540616; 4AT Bellelli 2014 PMID 25028342; MIND-USA Girard 2018 PMID 30346242)

psychiatryacutesyndromeadultgeriatricpediatricacuteinpatientoutpatienttransition
Start encounter →Print handoutPRODUCTION

Delirium dossier — DSM-5-TR 2022 + SCCM PADIS 2018 (Devlin PMID 30113379) + NICE CG103 2010 reaffirmed 2024 + CAM (Inouye 2006 PMID 16540616) + 4AT (Bellelli 2014 PMID 25028342) + CAM-ICU (Ely 2001 PMID 11730446) + RASS (Sessler 2002 PMID 12421743) + HELP (Inouye 1999 PMID 10089183) + MIND-USA (Girard 2018 PMID 30346242) + Marcantonio 2017 PMID 29020579 + CAPD (Traube 2014 PMID 24717459) + AGS Beers 2023 PMID 36968894 Covers F05.x DSM-5-TR delirium + F05.1 delirium superimposed on dementia + F10.231 alcohol withdrawal delirium + F13.231 benzodiazepine withdrawal delirium; prevalence ~ 30 % hospitalised older adults / ~ 40-80 % ICU / ~ 15-40 % post-op > 65 yr / ~ 25-30 % PICU; subtypes hyperactive ~ 25 % / hypoactive ~ 50 % (MOST commonly missed) / mixed ~ 25 % Treatment hierarchy: (1) identify + treat underlying cause (DELIRIUMS mnemonic — always first); (2) non-pharmacologic HELP / NICE bundle FIRST; (3) antipsychotic SECOND-LINE for severe agitation with safety risk ONLY — haloperidol 0.25-0.5 mg PO/IV geriatric (1-2 mg adult), risperidone 0.25-0.5 mg PO BID, quetiapine 12.5-25 mg PO QHS (Parkinson / Lewy body preferred), olanzapine 2.5-5 mg PO; (4) ICU PADIS A2F bundle (dexmedetomidine over benzodiazepine continuous infusion; Reade SPICE-III NEJM 2019 NEEDS_SOURCE_REVIEW); (5) alcohol/benzo withdrawal-delirium exception (lorazepam 1-2 mg IV q15-30 min titrated to CIWA-Ar/RASS; routes to psych.alcohol_withdrawal.core.v1; thiamine 100 mg IV BEFORE glucose) NO antipsychotic prophylaxis — MIND-USA Girard NEJM 2018 PMID 30346242 NEGATIVE for haloperidol + ziprasidone in ICU + Hope-ICU Page Lancet Resp Med 2013 NEEDS_SOURCE_REVIEW NEGATIVE for haloperidol prophylaxis AVOID benzodiazepines (worsen delirium; AGS Beers 2023 PMID 36968894) EXCEPT alcohol/benzo withdrawal-delirium where they are FIRST-LINE; AVOID anticholinergics (diphenhydramine, scopolamine, oxybutynin, TCAs; AGS Beers 2023); AVOID haloperidol in Parkinson / Lewy body / Parkinson-plus (use quetiapine); AVOID IM olanzapine + IM benzodiazepine combination (FDA olanzapine IM label cardiopulmonary depression); AVOID opioid excess (PADIS 2018 PMID 30113379); minimum effective dose for shortest duration for any antipsychotic; taper antipsychotic within 24-48 h of agitation resolution (NICE CG103 2010; AGS Beers 2023) CAM (Inouye 2006 PMID 16540616) sensitivity 94 % specificity 89 %; 4AT (Bellelli 2014 PMID 25028342) sensitivity 88 % specificity 88 %; CAM-ICU (Ely 2001 PMID 11730446) for non-verbal / ventilated; RASS (Sessler 2002 PMID 12421743) for sedation titration + CAM-ICU prerequisite (RASS ≥ -3 required); CAPD (Traube 2014 PMID 24717459) sensitivity 94 % specificity 79 % pediatric; p-CAM-ICU for verbal children HELP / NICE non-pharm bundle reduces delirium incidence ~ 37 % (Inouye NEJM 1999 PMID 10089183): reorientation + sleep hygiene + hydration + sensory aids (glasses, hearing aids, dentures) + early mobilisation + family presence + minimise restraints/lines/catheters + pain control without oversedation ICU PADIS A2F bundle (Devlin Crit Care Med 2018 PMID 30113379): Assess pain (CPOT/BPS); Both SAT (spontaneous awakening trial) + SBT (spontaneous breathing trial); Choice of analgesia (opioid-sparing multimodal); Delirium monitor (CAM-ICU + RASS q-shift); Early mobility (PT/OT daily); Family presence (flexible visiting hours) Post-discharge follow-up: brief neurocognitive screen (MoCA / MMSE) at 4-6 wk + 3 mo + 6 mo; 30-50 % have persistent features; 40-60 % progress to dementia within 1 yr; ~ 2-3× mortality at hospital + 1-yr; LOS prolongation; functional decline; caregiver burden (Marcantonio NEJM 2017 PMID 29020579) CRITICAL anti-patterns: NEVER use antipsychotics prophylactically (MIND-USA 2018 PMID 30346242 NEGATIVE; Hope-ICU 2013 NEGATIVE); NEVER use benzodiazepines in delirium EXCEPT alcohol/benzo withdrawal-delirium; NEVER give glucose before thiamine in alcohol-use disorder / Wernicke risk; NEVER skip cause-workup (DELIRIUMS mnemonic) — palliative treatment without cause-directed care leaves curable disease untreated; NEVER continue antipsychotic at discharge unless independent indication (AGS Beers 2023 PMID 36968894); NEVER omit ECG QTc monitoring with haloperidol or any antipsychotic; NEVER use haloperidol in Parkinson / Lewy body / Parkinson-plus syndromes (use quetiapine); NEVER co-administer IM olanzapine + IM benzodiazepine (FDA olanzapine IM label cardiopulmonary depression); NEVER omit medication reconciliation with anticholinergic burden score; NEVER omit CAM/4AT q-shift in older adult admission; NEVER mistake hypoactive delirium for depression or "natural ageing" — most clinically dangerous subtype (Marcantonio NEJM 2017 PMID 29020579); NEVER continue benzodiazepine continuous infusion in ICU when dexmedetomidine alternative available (PADIS 2018 PMID 30113379); NEVER omit post-discharge cognitive follow-up — 30-50 % have persistent features Severity triggers (10): hyperactive_delirium_with_safety_risk (severe — 1:1 sitter + non-pharm + low-dose haloperidol; routes to psych.suicidality.ed if SI emerges), hypoactive_delirium_missed_diagnosis (severe — high index of suspicion in every hospitalised older adult; CAM/4AT q-shift; DELIRIUMS systematic), delirium_with_alcohol_or_benzo_withdrawal (life-threatening — benzodiazepine FIRST-LINE exception + thiamine BEFORE glucose; routes to psych.alcohol_withdrawal.core.v1), delirium_with_anticholinergic_burden (severe — STOPP/Beers + ACB/DBI + deprescribe + pharmacy consult), post_operative_delirium (severe — multifactorial; opioid-sparing + early mobilisation + sleep protocol + family presence; HELP bundle adapted), icu_delirium_with_padis_bundle (severe — full A2F bundle; dexmedetomidine over benzo; CAM-ICU + RASS q-shift; NO prophylactic antipsychotic), delirium_due_to_infection_underlying (severe — workup + empirics within 1 h if sepsis features; routes to id.sepsis or id.geriatric-infection-syndromes), delirium_with_persistent_features_after_discharge (severe — 30-50 % persistent; MoCA/MMSE at 4-6 wk + 3 mo + 6 mo; dementia workup if persistent), end_of_life_terminal_delirium (moderate — palliative care + goals-of-care + symptom-targeted comfort; palliative sedation exception to no-benzo rule), delirium_in_pediatric_setting (moderate — CAPD + p-CAM-ICU; same hierarchy + family presence emphasised; pediatric weight-based antipsychotic with specialty consultation) Four setting playbooks: outpatient (detection + cause screen + medication review + caregiver education + post-discharge cognitive follow-up), ED (cause workup + stabilise immediate threats + non-pharm bundle + antipsychotic only for safety + thiamine before glucose if applicable + route to inpatient or cause-specific dossier), inpatient (daily HELP/NICE non-pharm bundle + cause workup completion + Beers/STOPP daily + antipsychotic taper + post-discharge cognitive follow-up planning), ICU (full PADIS A2F bundle + dexmedetomidine over benzo + CAM-ICU + RASS q-shift + opioid-sparing + family presence + daily SAT+SBT + early mobilisation) Action plan green/yellow/red includes early-detection caregiver education (new confusion, restlessness or withdrawal, sleep-wake disruption, hallucinations) + medication review (new meds in last 1-2 wk; anticholinergics; sleep aids) + reorientation + sensory aids + family presence + emergency triggers (severe agitation, withdrawal signs, SI emergence, stroke symptoms, seizure, inability to manage at home) Sibling differentiation: id.geriatric-infection-syndromes.v1 (delirium-as-sepsis-presentation; Loeb SNF UTI; compose), psych.alcohol_withdrawal.core.v1 (benzodiazepine FIRST-LINE exception + thiamine before glucose; compose), psych.opioid_use_disorder.core.v1 (intoxication causes hypoactive presentation; naloxone reversal; MAT longitudinal; AVOID benzo+opioid), id.sepsis.core.v1 (hour-1 bundle priority; compose), id.hsv-neonatal.core.v1 (neonate < 28 d out-of-scope; route), endo.dka.core.v1 (DKA-delirium; insulin+fluid+electrolyte priority; compose), psych.depression.core.v1 (depressive pseudo-dementia is the most important hypoactive delirium differential — CAM negative if pure depression), psych.suicidality.ed.core.v1 (SI emerging during agitation; compose not replace; handoff via action_plan red zone) Front-end note: no existing delirium-specific panel surface in src/components/panels/ today; this dossier is the back-end pathway pack. CAM / 4AT / CAM-ICU / RASS / CAPD / p-CAM-ICU / DRS-R-98 / ACB-score / Drug-Burden-Index panel surfaces flagged for future wire-up batch Workup IDs all resolve to registered umbrellas in clinical-tools-registry — workup.delirium_screen + workup.severe_agitation + workup.suicide_risk + workup.chest_pain + workup.insomnia Calculator IDs all resolve to registered tools — calc.phq9 + calc.gad7 + calc.audit_c + calc.ckd_epi_2021. calc.cam + calc.4at + calc.cam_icu + calc.rass + calc.capd + calc.p_cam_icu + calc.drs_r_98 + calc.acb_score + calc.drug_burden_index + calc.cssrs flagged for future clinical-tools-registry batch — referenced via narrative + workup.delirium_screen + workup.suicide_risk umbrellas for now Phenotype matrix (subtype × cause × age × baseline-cognitive × setting × severity × treatment-response) is encoded indirectly via severity_triggers + setting_playbooks + sibling_differentiation. First-class TS field for phenotype matrix is schema-blocked — see docs/framework-audit/shard-5-obped-id-state.md Schema-blocked queue Bayesian linkage (CAM positive LR+ ~ 9 sensitivity 94 % specificity 89 % per Inouye NEJM 2006 PMID 16540616; 4AT ≥ 4 LR+ ~ 8 sensitivity 88 % specificity 88 % per Bellelli 2014 PMID 25028342; CAM-ICU positive LR+ ~ 6-9 per Ely 2001 PMID 11730446; CAPD ≥ 9 LR+ ~ 5-7 per Traube 2014 PMID 24717459; age ≥ 65 with acute change LR+ ~ 3-5; pre-existing dementia LR+ ~ 4-5 incident delirium; ACB score ≥ 3 LR+ ~ 3-5; polypharmacy ≥ 10 LR+ ~ 2-4; DELIRIUMS-cause findings — Drugs LR+ ~ 5, Electrolytes LR+ ~ 5-10, Infection LR+ ~ 5, Intracranial LR+ ~ 20, Substance withdrawal LR+ ~ 10-20; T_screen = every hospitalised older adult ≥ 65 + every ICU + every post-op; T_workup = CAM/4AT positive triggers DELIRIUMS systematic; T_non_pharm_first = every confirmed delirium; T_antipsychotic = severe agitation with safety risk ONLY; T_benzodiazepine = alcohol/benzo withdrawal-delirium ONLY; T_admit = new + cause pending OR safety concern OR caregiver inability OR severe agitation; T_ICU = intubation/sedation/hemodynamic instability OR severe withdrawal-delirium; cross-dossier routing to psych.alcohol_withdrawal + psych.opioid_use_disorder + id.sepsis + id.geriatric-infection-syndromes + id.hsv-neonatal + endo.dka + psych.depression + psych.suicidality.ed documented in co-located research bundle). ROS/DDx LR seed data audited by npm run audit:ros-ddx-coverage (cross-cutting; not touched by this shard) PRODUCTION blockers / open gaps: (1) RxCUIs not yet validated via scripts/research/rxnav-validate.ts; haloperidol 5093 + risperidone 35636 + quetiapine 51272 + olanzapine 61381 + lorazepam 6470 + chlordiazepoxide 2598 + diazepam 3322 + thiamine 10632 + dexmedetomidine 48937 + propofol 6960 + guanfacine 5487 + melatonin 14635 PARTIAL; (2) manifest reuses psych.depression.core.v1.ts pointer per psych.adhd / psych.bipolar / psych.alcohol_withdrawal / psych.suicidality / psych.anxiety-disorders / psych.ptsd / psych.eating-disorders precedent — dedicated psych.delirium.core.v1.ts manifest out-of-shard-scope; (3) calc.cam + calc.4at + calc.cam_icu + calc.rass + calc.capd + calc.p_cam_icu + calc.drs_r_98 + calc.acb_score + calc.drug_burden_index + calc.cssrs not in clinical-tools-registry — referenced via workup.delirium_screen + workup.suicide_risk + narrative; (4) Hope-ICU Page Lancet Resp Med 2013 PMID NEEDS_SOURCE_REVIEW + Reade SPICE-III NEJM 2019 PMID NEEDS_SOURCE_REVIEW + STOPP/START v3 2023 PMID NEEDS_SOURCE_REVIEW + Trzepacz DRS-R-98 2001 PMID NEEDS_SOURCE_REVIEW + Norman J Am Geriatr Soc 2000 PMID NEEDS_SOURCE_REVIEW + Loeb SNF UTI 2001 PMID NEEDS_SOURCE_REVIEW + Sullivan CIWA-Ar 1989 PMID NEEDS_SOURCE_REVIEW + ASAM 2020 PMID NEEDS_SOURCE_REVIEW + Mayo-Smith PMID NEEDS_SOURCE_REVIEW + EAPC palliative delirium PMID NEEDS_SOURCE_REVIEW + ASA/ESA peri-op delirium PMID NEEDS_SOURCE_REVIEW + Inouye 1996 predictive model PMID NEEDS_SOURCE_REVIEW + Stanley-Brown 2012 SPI PMID NEEDS_SOURCE_REVIEW + Mann JAMA 2005 lethal-means PMID NEEDS_SOURCE_REVIEW — referenced by label only; not added to evidence.pmids per verification rule; (5) targeted test file pending (relies on dossier-contract.test.ts); (6) panel→dossier wire (CAM/4AT/CAM-ICU/RASS/CAPD panel) not authored — flagged for future wire-up batch; (7) neonatal delirium (< 28 d) referred to id.hsv-neonatal.core.v1 — out-of-shard-scope for this dossier; (8) complex neurodevelopmental delirium phenotypes (autism + delirium; intellectual disability + delirium) deferred to future pediatric dossiers

Entry points (9)

  • symptom
    Acute change in cognition / inattention / fluctuating mental status / altered level of consciousness in any care setting (DSM-5-TR 2022 criterion A+B; CAM Inouye NEJM 2006 PMID 16540616)
    acute_change_in_mental_status
  • symptom
    New agitation OR new withdrawn / drowsy / low-arousal state in hospitalised older adult — hyperactive ~25 % / hypoactive ~50 % / mixed ~25 % (Marcantonio NEJM 2017 PMID 29020579)
    agitation_or_hypoactivity_in_older_adult
  • lab_abnormality
    Positive CAM (feature 1 + 2 AND 3 OR 4; Inouye NEJM 2006 PMID 16540616) OR 4AT ≥ 4 (Bellelli Age Ageing 2014 PMID 25028342) on routine bedside screen
    positive_cam_or_4at_screen
  • lab_abnormality
    Positive CAM-ICU + RASS ≥ -3 in ventilated / non-verbal patient (Ely JAMA 2001 PMID 11730446; Sessler AJRCCM 2002 PMID 12421743)
    positive_cam_icu_screen
  • lab_abnormality
    Positive CAPD ≥ 9 (Traube Crit Care Med 2014 PMID 24717459) OR p-CAM-ICU in PICU patient
    positive_capd_or_p_cam_icu_pediatric
  • history
    Hospital admission of older adult ≥ 65 with risk factors (dementia, severe illness, prior delirium, visual/hearing impairment, dehydration, polypharmacy, alcohol use) — pre-emptive HELP bundle (Inouye NEJM 1999 PMID 10089183)
    hospital_admission_older_adult_with_risk_factors
  • history
    Post-operative > 65 yr — incidence 15-40 %; multifactorial pain + meds + sleep + electrolytes (Marcantonio NEJM 2017 PMID 29020579)
    post_operative_older_adult
  • history
    ICU admission ± mechanical ventilation — incidence 40-80 %; PADIS A2F bundle (SCCM PADIS 2018 Devlin PMID 30113379)
    icu_admission_with_mechanical_ventilation
  • history
    Alcohol or benzodiazepine cessation in last 72-96 h — withdrawal-delirium risk (CIWA-Ar; routes to psych.alcohol_withdrawal.core.v1)
    alcohol_or_benzodiazepine_cessation

Required inputs (28)

  • agerequired
    demographic • used at CONTEXT
    Drug + bundle selection; pediatric (CAPD + p-CAM-ICU; weight-based haloperidol/risperidone if used; Traube 2014 PMID 24717459); adult; geriatric ≥ 65 (highest prevalence; HELP bundle; AGS Beers PIM screen; lowest antipsychotic dose if used)
  • cognitive_baselinerequired
    history • used at CONTEXT
    Cognitive baseline (intact / MCI / mild-mod dementia / severe dementia) anchors change-from-baseline detection — central to delirium diagnosis vs underlying dementia (Inouye NEJM 2006 PMID 16540616; Marcantonio NEJM 2017 PMID 29020579)
  • temperaturerequired
    vital • used at INITIAL_WORKUP
    Fever / hypothermia — infection-as-cause workup; blunted fever response in frail older adult does not rule out infection (Norman J Am Geriatr Soc 2000 NEEDS_SOURCE_REVIEW)
  • heart_raterequired
    vital • used at INITIAL_WORKUP
    Tachycardia in sepsis / withdrawal / hyperthyroidism / dehydration — cause workup (Marcantonio NEJM 2017 PMID 29020579)
  • blood_pressurerequired
    vital • used at INITIAL_WORKUP
    Hypotension in sepsis / hemodynamic instability; hypertension in intracranial event / withdrawal (NICE CG103 2010; SSC 2026)
  • respiratory_raterequired
    vital • used at INITIAL_WORKUP
    Tachypnea in sepsis / hypoxia / acidosis (NICE CG103 2010)
  • spo2required
    vital • used at INITIAL_WORKUP
    Hypoxia as common cause + delirium precipitant (NICE CG103 2010; Marcantonio NEJM 2017 PMID 29020579)
  • glucoserequired
    lab • used at INITIAL_WORKUP
    Hypo- AND hyperglycaemia precipitate delirium; bedside fingerstick before any opioid / sedative / thiamine (NICE CG103 2010)
  • cam_or_4at_screenrequired
    symptom • used at ENTRY
    CAM (Inouye 2006 PMID 16540616) OR 4AT (Bellelli 2014 PMID 25028342) bedside delirium screen at admission + q-shift — sensitivity 88-94 % / specificity 88-89 %; mandates DELIRIUMS-cause workup if positive
  • rass_score
    symptom • used at CONTEXT
    Richmond Agitation-Sedation Scale (Sessler AJRCCM 2002 PMID 12421743) -5 unarousable to +4 combative; RASS ≥ -3 required to administer CAM-ICU; target -2 to 0 in ICU (PADIS 2018 PMID 30113379)
  • cam_icu_screen
    symptom • used at ENTRY
    CAM-ICU (Ely JAMA 2001 PMID 11730446) for non-verbal / intubated patient; combined with RASS; PADIS 2018 q-shift
  • capd_or_p_cam_icu
    symptom • used at ENTRY
    CAPD (Traube Crit Care Med 2014 PMID 24717459) for pre-verbal / non-verbal children in PICU; p-CAM-ICU for verbal children
  • cause_workup_deliriumsrequired
    history • used at DIFFERENTIAL
    DELIRIUMS mnemonic — Drugs / Electrolytes / Lack-of-meds / Infection / Reduced-sensory-input / Intracranial / Urinary-retention / Metabolic / Substances — every delirium needs underlying cause identified (Marcantonio NEJM 2017 PMID 29020579; NICE CG103 2010)
  • current_meds_with_anticholinergic_burdenrequired
    medication • used at CONTEXT
    Anticholinergic Burden (ACB) score or Drug Burden Index — diphenhydramine, scopolamine, oxybutynin, TCAs, antihistamines, antipsychotics, opioids, benzodiazepines all delirogenic in older adult (AGS Beers 2023 PMID 36968894; STOPP/START v3 2023 NEEDS_SOURCE_REVIEW)
  • alcohol_use_historyrequired
    history • used at CONTEXT
    Alcohol use + cessation timing — CIWA-Ar mandatory if last drink 72-96 h; routes to psych.alcohol_withdrawal.core.v1 for benzodiazepine-titrated regimen
  • benzodiazepine_or_opioid_use_historyrequired
    history • used at CONTEXT
    Benzodiazepine cessation → withdrawal-delirium; opioid intoxication / withdrawal → opioid-targeted workup; routes to psych.opioid_use_disorder.core.v1 if applicable
  • recent_surgery_or_anaesthesia_exposure
    history • used at CONTEXT
    Post-op delirium incidence 15-40 % > 65 yr; multifactorial pain + meds + sleep + electrolytes; HELP bundle; minimise benzodiazepine + opioid (Marcantonio NEJM 2017 PMID 29020579)
  • goals_of_care
    history • used at CONTEXT
    Goals-of-care alignment for terminal delirium (palliative); for ICU-vs-ward / DNR / family-presence decisions (NICE CG103 2010; EAPC palliative)
  • cbcrequired
    lab • used at INITIAL_WORKUP
    Leukocytosis / leukopenia in infection; blunted leukocytosis in frail older adult (Norman 2000 NEEDS_SOURCE_REVIEW)
  • bmprequired
    lab • used at INITIAL_WORKUP
    Na, K, Ca, glucose, BUN/Cr — electrolyte derangements are common delirium causes (Marcantonio NEJM 2017 PMID 29020579)
  • magnesiumrequired
    lab • used at INITIAL_WORKUP
    Hypo- AND hypermagnesaemia precipitate delirium; mandatory in withdrawal-delirium workup
  • lft
    lab • used at INITIAL_WORKUP
    Hepatic encephalopathy as cause; ammonia if cirrhosis (NICE CG103 2010)
  • tsh
    lab • used at INITIAL_WORKUP
    Hypo- AND hyperthyroidism precipitate delirium; baseline screen reasonable in initial workup (NICE CG103 2010)
  • b12
    lab • used at BRANCHING_WORKUP
    B12 deficiency mimics delirium in older adult / restricted diet (NICE CG103 2010)
  • urinalysisrequired
    lab • used at INITIAL_WORKUP
    UTI as common cause in older adult; UA + UC; Loeb criteria for SNF UTI (id.geriatric-infection-syndromes sibling)
  • cxrrequired
    imaging • used at INITIAL_WORKUP
    Pneumonia as common cause; CXR mandatory in older adult delirium workup (NICE CG103 2010)
  • ecg
    imaging • used at INITIAL_WORKUP
    QTc baseline before haloperidol / antipsychotic; arrhythmia or recent MI may cause delirium (DailyMed haloperidol; AGS Beers 2023)
  • ct_head
    imaging • used at BRANCHING_WORKUP
    Intracranial bleed / stroke / mass — CT head if focal signs, anticoagulation, fall + head strike, no clear cause, or hyperactive delirium without obvious cause (NICE CG103 2010)

12-phase flow (12)

  1. 1FRAME
    Confirm DSM-5-TR 2022 delirium criteria (A: disturbance of attention + awareness; B: develops over short time + fluctuates; C: additional cognitive disturbance; D: not better explained by pre-existing/established neurocognitive disorder + not in context of severely reduced arousal/coma; E: evidence of medical cause / intoxication / withdrawal / multiple etiologies). CAM 4-feature bedside screen anchor (Inouye NEJM 2006 PMID 16540616)
    advance: DSM-5-TR criteria met + change-from-baseline cognition documented + dementia / primary psychiatric diagnosis ruled out as sole explanation
  2. 2ENTRY
    Trigger from acute mental status change OR positive CAM / 4AT / CAM-ICU / CAPD / p-CAM-ICU screen OR hospital admission of older adult with risk factors (NICE CG103 2010; Marcantonio NEJM 2017 PMID 29020579)
    inputs: age, cam_or_4at_screen
    advance: Entry criteria documented; subtype (hyperactive / hypoactive / mixed) provisionally assigned
  3. 3CONTEXT
    Baseline cognition (intact / MCI / mild-mod dementia / severe dementia); pre-admission medication list with anticholinergic burden (ACB / DBI); alcohol / benzo / opioid use; recent surgery / anaesthesia exposure; setting (ED / ward / ICU / post-op / palliative / peds / SNF); goals of care
    inputs: cognitive_baseline, current_meds_with_anticholinergic_burden, alcohol_use_history, benzodiazepine_or_opioid_use_history, recent_surgery_or_anaesthesia_exposure
    advance: Personalisation data captured; baseline cognition + medication review + substance use + goals of care documented
  4. 4RED_FLAGS
    Hyperactive delirium with safety risk (falls, self-extubation, aggression — 1:1 sitter); life-threatening cause (alcohol/benzo withdrawal-delirium → CIWA-Ar + benzo per protocol → psych.alcohol_withdrawal.core.v1; sepsis-delirium → hour-1 bundle → id.sepsis.core.v1; DKA-delirium → endo.dka.core.v1; uremic delirium; encephalitis; hypoglycaemia [fingerstick first!]; opioid intoxication / withdrawal; hyper/hyponatremia; non-convulsive status epilepticus; intracranial bleed); active SI emerging during agitation → routes to psych.suicidality.ed.core.v1
    inputs: temperature, heart_rate, blood_pressure, spo2, glucose
    advance: Immediate safety threats stabilised; bedside glucose checked; thiamine 100 mg IV given BEFORE glucose if Wernicke / alcohol-withdrawal suspected; 1:1 sitter assigned if severe agitation; cause-specific dossier route if life-threatening cause
  5. 5INITIAL_WORKUP
    CAM / 4AT at triage; full vitals (T, HR, BP, RR, SpO2, glucose); chemistry (Na, K, Ca, Mg, glucose, BUN/Cr, LFT); CBC; UA + UC; CXR; ECG; medication reconciliation with anticholinergic burden score (ACB / Drug Burden Index); pulse oximetry; review STOPP/Beers PIM list (NICE CG103 2010; Marcantonio NEJM 2017 PMID 29020579; AGS Beers 2023 PMID 36968894)
    inputs: cbc, bmp, magnesium, urinalysis, cxr, temperature, heart_rate, blood_pressure
    advance: Baseline workup returned; DELIRIUMS-cause categories systematically reviewed
  6. 6BRANCHING_WORKUP
    CT head if focal signs / anticoagulation / fall + head strike / no clear cause; LP if meningitis / encephalitis / fever + delirium without clear source; EEG if non-convulsive status epilepticus suspected (LOC fluctuation without clear cause); TSH / B12 / ammonia (hepatic encephalopathy) / cortisol / heavy metals / drug levels (digoxin, lithium, valproate); blood cultures if febrile; toxicology screen if suspected ingestion / withdrawal (NICE CG103 2010; Marcantonio NEJM 2017 PMID 29020579)
    advance: Targeted workup obtained when triggered; cause identified or multifactorial assigned
  7. 7DIFFERENTIAL
    Delirium (DSM-5-TR) vs major neurocognitive disorder (dementia — chronic; no acute change; no fluctuation; attention often preserved early) vs depressive pseudo-dementia (slow tempo; PHQ-9 high; routes to psych.depression.core.v1) vs psychotic disorder (younger; chronic; preserved attention) vs non-convulsive status epilepticus (EEG diagnostic) vs Wernicke encephalopathy (triad ophthalmoplegia + ataxia + confusion; thiamine 100 mg IV BEFORE glucose) vs CNS infection vs intracranial pathology vs hypoglycaemia vs metabolic derangement vs intoxication/withdrawal (Marcantonio NEJM 2017 PMID 29020579; NICE CG103 2010)
    inputs: cause_workup_deliriums
    advance: Working diagnosis + cause(s) + subtype assigned
  8. 8RISK_STRATIFICATION
    Subtype (hyperactive ~25 % / hypoactive ~50 % / mixed ~25 %); cause identified vs multifactorial; severity (DRS-R-98 or delirium index — NEEDS_SOURCE_REVIEW); duration (days; > 1 wk worsens outcomes); setting (ICU vs ward); functional / cognitive baseline; outcome predictors (mortality 2-3× hospital + 1-yr; LOS; dementia trajectory; functional decline) (Marcantonio NEJM 2017 PMID 29020579)
    inputs: cam_or_4at_screen, cognitive_baseline
    advance: Subtype + cause + severity + setting + safety risk documented
  9. 9TREATMENT
    Hierarchy: (1) IDENTIFY + TREAT UNDERLYING CAUSE (DELIRIUMS mnemonic — always first). (2) NON-PHARMACOLOGIC FIRST — HELP / NICE bundle: reorientation (clock, calendar, name); sleep hygiene (dim lights at night, no overnight vitals if possible, daytime light + activity); hydration / nutrition; restore sensory aids (glasses, hearing aids, dentures); early mobilisation; family presence / familiar objects; minimise restraints; minimise lines / catheters; pain control without oversedation (Inouye NEJM 1999 PMID 10089183; NICE CG103 2010). (3) PHARMACOLOGIC ONLY for severe agitation with safety risk — haloperidol 0.25-0.5 mg PO/IV (geriatric; QTc monitor; AVOID Parkinson / Lewy body); risperidone 0.25-0.5 mg PO BID; quetiapine 12.5-25 mg PO QHS (preferred Parkinson / Lewy body); olanzapine 2.5-5 mg PO; AVOID benzodiazepines EXCEPT in alcohol/benzo withdrawal-delirium (lorazepam 1-2 mg IV q15-30 min titrated to CIWA-Ar / RASS — routes to psych.alcohol_withdrawal.core.v1). NO antipsychotic prophylaxis (MIND-USA Girard NEJM 2018 PMID 30346242 NEGATIVE for haloperidol + ziprasidone; Hope-ICU Page Lancet Resp Med 2013 NEEDS_SOURCE_REVIEW). SCCM PADIS 2018 A2F bundle for ICU (Assess pain + Both SAT/SBT + Choice of analgesia + Delirium monitor + Early mobility + Family presence; Devlin PMID 30113379)
    inputs: current_meds_with_anticholinergic_burden
    advance: Underlying cause(s) being treated + HELP/NICE non-pharm bundle initiated + antipsychotic prescribed ONLY if severe agitation with safety risk + benzodiazepine reserved for withdrawal-delirium only + ICU PADIS A2F bundle in place if ICU setting
  10. 10DISPOSITION
    ED → admit ward if new delirium + cause workup pending OR severity requires monitoring; ICU if intubation / sedation needed OR hemodynamic instability OR severe alcohol/benzo withdrawal with delirium tremens; SNF / home with caregiver if mild + cause identified + treated + safe environment; palliative for terminal delirium with goals-of-care alignment; post-discharge cognitive follow-up for persistent features (NICE CG103 2010)
    inputs: goals_of_care
    advance: Level of care set; goals-of-care discussed if older adult / advanced illness
  11. 11MONITORING
    CAM / 4AT q-shift (ward) or q-2-hr (ICU CAM-ICU + RASS continuous + CPOT pain assessment); RASS target -2 to 0 for ICU; SAT (spontaneous awakening trial) + SBT (spontaneous breathing trial) daily per PADIS 2018; medication reconciliation daily; non-pharm bundle adherence; antipsychotic taper as soon as agitation resolves (typically 24-48 h post-resolution); duration tracking (mortality + LOS predictor); pediatric CAPD + p-CAM-ICU q-shift (Traube 2014 PMID 24717459)
    advance: Delirium resolving OR persistent OR worsening; antipsychotic tapered when agitation resolves
  12. 12FOLLOWUP
    Post-discharge cognitive assessment 4-6 wk + 3 mo + 6 mo (persistent features in 30-50 %); caregiver support / education; may unmask dementia (40-60 % of survivors progress within 1 yr; Marcantonio NEJM 2017 PMID 29020579); functional rehabilitation referral; primary-care + geriatrics follow-up; advance care planning revisit
    advance: Post-discharge cognitive follow-up scheduled + caregiver education complete + relevant referrals placed