Part 3: Delirium Assessment, Prevention, and Management

Delirium in the ICU: Overview

Delirium is an acute, fluctuating disturbance in attention and awareness, accompanied by additional cognitive dysfunction (disorientation, memory deficit, language disturbance, visuospatial impairment, or perceptual disturbance) that is not better explained by a pre-existing neurocognitive disorder and does not occur in the context of a severely reduced level of arousal (e.g., coma).^{1 2}

Epidemiology

Consequences of ICU Delirium

Delirium Subtypes

Clinical Pearl: Hypoactive delirium is the most common subtype and the most frequently missed. It is associated with worse outcomes than hyperactive delirium. Routine screening with validated tools is essential to detect it.²

Delirium Assessment Tools

The Confusion Assessment Method for the ICU (CAM-ICU) — Complete Flowsheet

The CAM-ICU is the most widely used and validated bedside delirium assessment tool for ICU patients. It can be completed in 1–2 minutes by trained nurses, physicians, or other providers. Sensitivity is 80–95% and specificity is 89–100% when performed by trained assessors.^{2 5}

Prerequisites before performing the CAM-ICU:

• The patient must have a RASS score of −3 or higher (responsive to at least voice)
• If RASS is −4 or −5 → patient is in coma → CAM-ICU cannot be assessed (record as “unable to assess — coma”)

CAM-ICU Flowsheet: Four Features

The CAM-ICU is POSITIVE (delirium present) if Feature 1 AND Feature 2 are present, PLUS either Feature 3 OR Feature 4.

Feature 1: Acute Onset or Fluctuating Course

Scoring: If YES to any of the above → Feature 1 is POSITIVE → proceed to Feature 2. If NO to all → Feature 1 is NEGATIVE → CAM-ICU is NEGATIVE (no delirium). STOP.

Feature 2: Inattention — Assessed Using the Attention Screening Examination (ASE)

Two methods are available; use whichever the patient can perform:

Method A — Auditory (Letters Test) (PREFERRED)

Read the following 10 letters slowly (one letter per second) in a monotone voice. Instruct the patient: “I am going to read you a series of 10 letters. Squeeze my hand whenever you hear the letter ‘A’.”

Letter sequence: S A V E A H A A R T

Errors = failing to squeeze on “A” OR squeezing on a non-“A” letter.

Method B — Visual (Pictures Test)

Show a set of 5 pictures. Then show a set of 10 pictures (the original 5 mixed with 5 new pictures). The patient must nod “yes” or “no” to indicate whether they have seen each picture before. Count errors (incorrect responses).

Scoring Feature 2:

• 0–2 errors → Feature 2 is NEGATIVE (attention intact) → CAM-ICU is NEGATIVE. STOP.
• > 2 errors → Feature 2 is POSITIVE (inattention present) → Proceed to Feature 3.

Feature 3: Altered Level of Consciousness

Scoring:

• RASS = 0 (alert and calm) → Feature 3 is NEGATIVE → Proceed to Feature 4
• RASS ≠ 0 (any score other than 0) → Feature 3 is POSITIVE → CAM-ICU is POSITIVE (delirium present). STOP.

Feature 4: Disorganized Thinking

Ask the following Yes/No questions AND perform a simple command:

Set A (use on odd calendar days or first assessment):

Set B (use on even calendar days or alternate assessments):

Command: “Hold up this many fingers” (examiner holds up 2 fingers). After the patient does this, say “Now do the same thing with the other hand” (the patient should hold up 2 fingers on the other hand without being shown again). If patient is unable to move both arms, substitute: “Add one more finger” (patient should hold up 3 fingers).

Scoring Feature 4:

• Count errors from the questions (each incorrect answer = 1 error) and the command (inability to complete = 1 error)
• 0–1 errors → Feature 4 is NEGATIVE → CAM-ICU is NEGATIVE
• > 1 error → Feature 4 is POSITIVE → CAM-ICU is POSITIVE (delirium present)

CAM-ICU Summary Flowsheet

Step 1: Assess RASS
  ├── RASS −4 or −5 → COMA → Cannot assess CAM-ICU → Reassess later
  └── RASS ≥ −3 → Proceed to CAM-ICU

Step 2: Feature 1 — Acute onset or fluctuating course?
  ├── NO → CAM-ICU NEGATIVE (No Delirium)
  └── YES → Proceed to Feature 2

Step 3: Feature 2 — Inattention (ASE Letters or Pictures)?
  ├── ≤ 2 errors → CAM-ICU NEGATIVE (No Delirium)
  └── > 2 errors → Proceed to Feature 3

Step 4: Feature 3 — Altered level of consciousness (RASS ≠ 0)?
  ├── YES (RASS ≠ 0) → CAM-ICU POSITIVE (Delirium Present)
  └── NO (RASS = 0) → Proceed to Feature 4

Step 5: Feature 4 — Disorganized thinking?
  ├── > 1 error → CAM-ICU POSITIVE (Delirium Present)
  └── ≤ 1 error → CAM-ICU NEGATIVE (No Delirium)

CAM-ICU Assessment Frequency

Intensive Care Delirium Screening Checklist (ICDSC) — Complete Scoring

The ICDSC is an 8-item checklist completed by the bedside nurse based on observations over an 8–24 hour period. Each item is scored 0 (absent) or 1 (present). Total score ranges from 0 to 8. A score ≥ 4 indicates delirium. Sensitivity is 74–99% and specificity is 64–82%.⁶

ICDSC — Complete Scoring Table

Choosing Between CAM-ICU and ICDSC

Recommendation: Both tools are acceptable. The CAM-ICU has higher specificity and is the most widely studied. The ICDSC may have higher sensitivity in some settings and captures subsyndromal delirium. An institution should select one tool and ensure consistent training and compliance.¹

Delirium Risk Factors

Non-Modifiable Risk Factors

Modifiable Risk Factors

Delirium Prevention

Non-Pharmacologic Delirium Prevention Bundle

Non-pharmacologic, multicomponent prevention strategies are the cornerstone of delirium prevention and are recommended for all ICU patients (strong recommendation).^{1 7}

Pharmacologic Delirium Prevention

The 2018 guidelines do NOT recommend routine pharmacologic prophylaxis for delirium in most ICU patients. Specific findings:¹

Exception — Perioperative dexmedetomidine: In cardiac and non-cardiac surgical populations, perioperative dexmedetomidine has shown delirium-preventive effects in some trials. This may be considered for high-risk surgical ICU patients (conditional recommendation based on limited evidence).⁸

Delirium Management

Step 1: Identify and Treat Underlying Causes

When delirium is identified, the first priority is to identify and treat reversible precipitating factors. The mnemonic “THINK” can be used:^{1 9}

Step 2: Non-Pharmacologic Management

All non-pharmacologic prevention strategies (see prevention bundle above) should be continued and intensified during active delirium.^{1 7}

Step 3: Pharmacologic Management of Delirium

The 2018 guidelines recommend AGAINST the routine use of haloperidol or atypical antipsychotics for the treatment of ICU delirium (conditional recommendation), based on the MIND-USA and AID-ICU trials which found no significant benefit on delirium duration, ventilator-free days, or mortality.^{1 10 11}

However, antipsychotics may be considered in the following circumstances:

• Patients with severe agitation that poses an immediate safety risk to the patient or staff
• Hyperactive delirium with distressing symptoms (hallucinations, severe agitation) refractory to non-pharmacologic measures
• As a bridge while treating reversible underlying causes
• Patients with delirium and significant psychological distress

Haloperidol

Quetiapine

Olanzapine

Dexmedetomidine for Delirium-Related Sedation

Dexmedetomidine is NOT an antipsychotic but is the preferred sedative for patients with delirium who require sedation, based on its lower delirium-promoting profile compared to benzodiazepines and its potential to facilitate delirium resolution.^{1 10 12}

Pharmacologic Treatment Algorithm for ICU Delirium

Step 1: Identify and treat underlying causes (THINK)
        ↓
Step 2: Implement/intensify non-pharmacologic bundle
        ↓
Step 3: Assess delirium subtype and severity
        ├── Hypoactive delirium (calm, inattentive)
        │   → Non-pharmacologic interventions are PRIMARY
        │   → Remove sedatives if possible
        │   → Optimize sleep (quetiapine 25–50 mg PO qHS may help with sleep cycle)
        │   → Pharmacologic treatment NOT routinely recommended
        │   → Do NOT use haloperidol (may worsen sedation)
        │
        ├── Hyperactive delirium (agitated, hallucinations, combative)
        │   → Non-pharmacologic interventions FIRST
        │   → If safety concern or significant distress:
        │       → Haloperidol 2–5 mg IV q 30 min–4 h PRN (check QTc)
        │       → OR quetiapine 25–50 mg PO/NG q 12 h if enteral route available
        │   → If on mechanical ventilation requiring sedation:
        │       → Dexmedetomidine preferred sedative
        │       → AVOID benzodiazepines (worsen delirium)
        │
        └── Mixed delirium
            → Combine approaches based on predominant symptoms
            → Reassess subtype at each assessment

Duration of Antipsychotic Therapy

• If antipsychotics are initiated for delirium, plan for taper and discontinuation
• Reassess need at least daily
• Once delirium resolves (CAM-ICU negative for ≥ 24 hours), taper and discontinue antipsychotics over 2–5 days
• Do NOT continue antipsychotics at ICU discharge unless the patient had a pre-existing psychiatric indication
• Document the indication, planned duration, and taper plan in the medical record

Benzodiazepines and Delirium

Benzodiazepines are a major risk factor for ICU delirium and should be avoided whenever possible.^{1 2}

When benzodiazepines are appropriate:

• Active seizures (first-line)
• Alcohol or sedative/hypnotic withdrawal (specific benzodiazepine protocols — see Part 5)
• Procedural anxiolysis/sedation (single dose)
• Refractory agitation not responding to non-benzodiazepine agents (rescue)
• Patients on chronic benzodiazepine therapy (continue at prior dose to avoid withdrawal)
• Anxiety disorders with documented benzodiazepine dependence

Medications That Increase Delirium Risk

The following medications should be reviewed and minimized in ICU patients at risk for delirium:¹

Delirium Documentation and Communication

Recommended Documentation Elements

Handoff Communication for Delirium

Include in every nurse-to-nurse and nurse-to-provider handoff:

1. Current delirium status (CAM-ICU/ICDSC result)
2. Trend over past 24–48 hours
3. Identified contributing factors
4. Current management plan
5. Outstanding workup (pending labs, imaging, medication review)

References