# Delirium Assessment in the ICU

Use a structured arousal-first approach: establish wakefulness with RASS, then screen assessable ICU patients serially with CAM-ICU or ICDSC. A positive screen should trigger medication, physiologic, infectious, neurologic, and environmental review rather than reflex antipsychotic treatment.

**Clinical question:** How should clinicians screen, interpret, and act on delirium assessments in critically ill adults?

Updated: 2026-09-15T18:23:51.781194+00:00

## What matters in practice
- Screen ICU patients with a validated instrument—CAM-ICU or ICDSC—rather than relying on unstructured bedside impression; either can be completed in approximately 2–5 minutes. [21]
- Assess arousal before delirium testing. A RASS of -3 or lower commonly renders patients unassessable with standard delirium instruments; document coma/deep sedation and reassess after arousal improves. [21][24]
- Interpret a positive delirium screen in the context of sedative exposure and arousal: delirium positivity is substantially more frequent at RASS -2 than at RASS -1 to 0. [21]
- A positive screen should prompt active reversal of modifiable precipitants—especially benzodiazepine exposure, immobility, restraint use, sleep disruption, sensory/environmental deprivation, and uncontrolled pain—while evaluating for medical and neurologic causes. [14][16]
- Do not use antipsychotics as routine prophylaxis in older critically ill adults; current SCCM guidance makes no recommendation for antipsychotics to treat delirium because of insufficient evidence. [5][7]

## Use an arousal-first delirium assessment sequence

Separate inability to assess from a negative delirium screen.

Begin each assessment by documenting level of arousal with the Richmond Agitation-Sedation Scale (RASS). CAM-ICU and ICDSC results are difficult to interpret in deep sedation or coma; patients at RASS -3 are commonly classified as unable to assess in delirium studies, whereas assessment data are more established from RASS 0 through -2. [21][24]

When the patient is sufficiently arousable, perform CAM-ICU or ICDSC rather than substituting routine orientation questions or a general impression of confusion. Both tools are validated for ICU use, including in mechanically ventilated nonverbal patients, and can be completed in 2–5 minutes. [21][22]

Record the paired result as arousal status plus delirium status—for example, “RASS -4, delirium unassessable” or “RASS -1, CAM-ICU positive.” This distinguishes coma/deep sedation from a negative screen and makes serial change interpretable during sedation adjustment and recovery. [21][24]
- RASS -3 or lower: document inability to complete a standard delirium assessment; reassess when arousal improves. [21][24]
- RASS -2 to 0: proceed with CAM-ICU or ICDSC, while recognizing that reduced arousal can increase delirium-screen positivity. [21]
- Positive screen: initiate an etiologic and iatrogenic review during the same clinical evaluation; do not treat the instrument result alone. [14][16]

*Arousal-dependent interpretation of ICU delirium testing. [21][24]*

| Bedside state | Assessment action | Interpretation and next step |
| --- | --- | --- |
| Deeply sedated or comatose; RASS -3 or lower | Do not label the result negative; document delirium assessment as unassessable and repeat after arousal improves. [21][24] | Standard delirium-test data are limited in this range; review whether sedation can be lightened safely. [21] |
| Arousable; RASS -2 to 0 | Perform CAM-ICU or ICDSC. [21] | A positive result supports delirium but must be interpreted with sedation level; positivity was 77% at RASS -2 versus 23% at RASS -1 to 0 in pooled assessment data. [21] |
| Agitated; RASS greater than +1 | Assess for immediate danger, pain, hypoxemia, device intolerance, withdrawal, and delirium; use the local agitation and delirium protocol. [1] | Once safety and arousal permit, complete a validated delirium screen and address precipitating causes. [1][21] |

## Choose CAM-ICU or ICDSC and preserve serial comparability

Use one trained workflow consistently within the unit.

CAM-ICU is a practical choice when a brief, repeated bedside determination is needed in verbal or nonverbal ICU patients. In its original ICU validation cohort, CAM-ICU showed high interrater reliability and sensitivity of 95% to 100% and specificity of 89% to 93% against a reference delirium assessment. [22]

ICDSC is also guideline-endorsed for ICU screening. Tool performance and reproducibility vary by ICU population, particularly in patients with neurologic injury, communication impairment, aphasia, sensory deficits, or neuroprotective sedation; in these settings, interpret an instrument result alongside the neurologic examination and the expected trajectory of the primary brain injury. [21][23]

Avoid switching instruments merely because a result is unexpected. A change from CAM-ICU to ICDSC changes the measurement method, not necessarily the patient’s mental state. Maintain the same tool for serial bedside monitoring when feasible, and escalate discordant or clinically implausible results to direct clinical evaluation rather than assuming either a false-positive or false-negative screen. [21][23]
- CAM-ICU: validated in ICU patients who may be nonverbal because of mechanical ventilation. [22]
- ICDSC: appropriate for rapid clinical observation, but cognitive and psychiatric items may be under-recognized in neurocritical care practice. [23]
- Neither tool replaces focused assessment for focal deficits, seizure, meningitis/encephalitis, or structural brain injury when the examination suggests those conditions. EEG may help confirm metabolic encephalopathies or infectious encephalitides with characteristic patterns. [14]

*Tool-selection considerations for routine ICU delirium screening. [21][22][23]*

| Tool | Useful setting | Key limitation affecting action |
| --- | --- | --- |
| CAM-ICU | Brief repeated screening, including patients unable to speak because of mechanical ventilation. [22] | Do not interpret as negative during coma/deep sedation; arousal level materially affects positivity. [21] |
| ICDSC | Routine nursing observation-based screening in the ICU. [21][23] | Neurocritical care deficits and communication limitations can reduce recognition of cognitive and psychiatric features. [23] |
| Edinburgh Delirium Test Box–ICU | Potential longitudinal attention assessment in selected arousable patients. [24] | Validation was small; a score of 5 or lower was 100% sensitive and 92% specific in the reported cohort, but it is not the routine guideline screening standard. [24] |

## Treat a positive screen as a trigger for cause-directed reassessment

Prioritize reversible iatrogenic and physiologic contributors before symptom-suppressing medication.

For a new positive CAM-ICU or ICDSC, first compare the result with the contemporaneous RASS and recent sedative changes. Benzodiazepines are an independent delirium risk factor in ICU patients; evaluate whether ongoing benzodiazepine exposure is necessary and whether the sedation target can be safely lightened. [14]

Review potentially modifiable ICU precipitants at the bedside: immobilization, physical restraints, isolation or absent visitation, lack of visible daylight, excessive noise or light, and disrupted sleep. Also reassess pain and analgesic strategy, because pain control and reduction of avoidable distress are integrated components of PADIS-based ICU care. [14][18]

Use the examination to decide whether delirium may instead reflect a neurologic emergency or a primary neurologic disorder. New focal deficits, unexplained depressed consciousness, or concern for seizure should prompt a neurologic assessment; EEG can be useful when evaluating metabolic encephalopathy or infectious encephalitis with characteristic EEG patterns. [14]
- Medication review: identify benzodiazepines and assess whether sedation depth is greater than clinically required. [14][21]
- Environment and mobility review: mobilize or rehabilitate when clinically feasible, minimize restraints, restore day-night cues, and facilitate family participation when possible. Multicomponent nonpharmacologic strategies have the strongest comparative evidence for reducing ICU delirium incidence. [5][9]
- Neurologic escalation: investigate focal findings, evolving consciousness changes, or seizure concern rather than attributing them to delirium alone. [14]

### Agitated delirium versus hypoactive delirium

Do not reserve screening for overt agitation. Delirium can present with reduced arousal, and CAM-ICU positivity increases as arousal declines within the assessable range. A quiet patient with RASS -2 may therefore warrant the same validated screen and etiologic review as an agitated patient. [21]

For agitation with RASS greater than +1, follow the unit’s agitation and delirium protocol while immediately evaluating pain, sedation, and acute physiologic distress. Once safety permits, obtain a structured delirium assessment rather than treating agitation as synonymous with delirium. [1]

*Positive delirium screen: bedside branches that change the next action. [1][14][18][21]*

| Finding | Immediate interpretation | Next action |
| --- | --- | --- |
| Positive screen after recent sedative escalation or RASS near -2 | Sedation depth may contribute to the observed cognitive abnormality. [21] | Review sedatives, reassess the sedation target, and repeat structured assessment after clinically appropriate arousal improvement. [14][21] |
| Positive screen with benzodiazepine exposure | Benzodiazepines are an independent ICU delirium risk factor. [14] | Reassess indication and minimize avoidable exposure while maintaining required seizure, withdrawal, or procedural management. [14] |
| Positive screen with restraint, immobility, isolation, or sleep/circadian disruption | Multiple modifiable environmental precipitants are present. [14] | Implement a multicomponent strategy emphasizing mobility/rehabilitation, orientation, environmental control, and family participation where feasible. [5][9] |
| Positive screen plus focal deficit, atypical course, or unexplained depressed consciousness | Consider structural, epileptic, infectious, or metabolic neurologic disease. [14] | Perform targeted neurologic evaluation; use EEG when metabolic encephalopathy or infectious encephalitis is suspected. [14] |

## Repeat screening across changing sedation and illness states

A single negative assessment does not establish delirium-free status throughout an ICU stay.

Repeat a validated assessment whenever the patient becomes newly arousable, after a meaningful change in sedation or neurologic status, and as part of the unit’s routine delirium-monitoring workflow. Delirium is dynamic, and screening studies and ICU protocols rely on serial rather than one-time assessments. [20][21][22]

In acute stroke, early serial testing is particularly important: delirium began on the first day in 67.3% of affected patients and within 5 days in all affected patients in one prospective cohort. CAM-ICU demonstrated 76% sensitivity and 98% specificity against DSM assessment in that poststroke population, but aphasia and stroke-related deficits still require clinical interpretation. [20]

Trend delirium status with RASS and exposure changes rather than treating either measure as isolated. A patient transitioning from RASS -4 to -1 has become assessable; a new positive CAM-ICU at that point may represent emerging delirium, sedation-related cognitive effects, or both, and should trigger reassessment of the full clinical context. [21]
- Document the tool used, RASS, result, and whether the patient was unassessable because of coma/deep sedation. [21][24]
- Re-screen after improvement in arousal rather than carrying forward a prior “unable to assess” designation. [21][24]
- In neurologic ICU patients, correlate serial screen changes with focal examination findings and the underlying neurologic diagnosis. [20][23]

*Events that should prompt repeat ICU delirium assessment. [20][21][22][24]*

| Clinical event | Why reassessment matters | Documentation target |
| --- | --- | --- |
| Arousal improves from deep sedation/coma | The patient may become eligible for a valid bedside delirium screen. [21][24] | RASS plus CAM-ICU or ICDSC result. |
| Sedative strategy or RASS changes | Arousal affects likelihood of delirium-screen positivity. [21] | Current RASS, drug exposure change, and delirium result. |
| New agitation or reduced engagement | Either hyperactive or hypoactive presentation may represent delirium. [1][21] | Safety assessment, RASS, validated screen, and precipitant review. |
| Early period after acute stroke | Most detected poststroke delirium occurred within 5 days in the reported cohort. [20] | Daily serial assessment interpreted with stroke deficits. |

## Link screening to nonpharmacologic care and avoid reflex antipsychotic use

Delirium detection is valuable only when it changes daily ICU decisions.

Use each positive assessment to activate a multicomponent nonpharmacologic plan: optimize orientation and cognition, mobilize early when feasible, assess and control pain, reduce avoidable environmental sleep disruption, and involve family when possible. Comparative evidence identifies multicomponent strategies as the most effective nonpharmacologic approach for reducing ICU delirium incidence; SCCM also conditionally recommends enhanced mobilization/rehabilitation over usual mobilization/rehabilitation. [4][5][9]

Do not prescribe an antipsychotic solely because CAM-ICU or ICDSC is positive. SCCM’s focused update was unable to recommend antipsychotics for treatment of delirium in critically ill adults, and SCCM guidance for older critically ill adults suggests against antipsychotics for delirium prevention. [5][7]

For patients who require sedation, choice of sedative should be considered as part of the delirium-assessment response rather than separately. SCCM conditionally recommends dexmedetomidine over propofol for sedation in critically ill adults; this does not convert a positive delirium screen into an indication for drug treatment, but it can inform sedation strategy when sedation remains necessary. [5]
- Use structured delirium results in daily sedation, mobility, pain, and sleep planning rather than as a stand-alone quality metric. [5][16]
- Avoid antipsychotic prophylaxis in older critically ill adults. [7]
- When sleep disruption is clinically relevant, SCCM conditionally recommends melatonin in critically ill adults; systematic-review findings suggest possible improvement in perceived sleep and reduced delirium, with uncertainty from bias and inconsistency. [5][19]

*Assessment-linked management decisions in ICU delirium. [5][7][9][19]*

| Assessment-linked problem | Preferred response | What not to infer |
| --- | --- | --- |
| Positive delirium screen with modifiable ICU precipitants | Use a multicomponent nonpharmacologic intervention with mobility/rehabilitation, environmental optimization, orientation, and family participation as feasible. [5][9] | A positive screen alone does not identify a specific drug-responsive delirium subtype. |
| Older critically ill adult at risk for delirium | Use a geriatric model of care when available and avoid antipsychotic prophylaxis. [7] | Antipsychotics should not be used routinely to prevent delirium. [7] |
| Sedation remains necessary after a positive screen | Reassess sedation target and consider SCCM’s conditional preference for dexmedetomidine over propofol. [5] | Sedative selection does not replace diagnostic evaluation for delirium precipitants. |
| Sleep disruption contributing to an ICU delirium-risk profile | Apply environmental sleep measures; melatonin is conditionally recommended by SCCM. [5][19] | Optimal melatonin dose and administration timing remain uncertain. [19] |

## References
1. PupillOmetry for preDIction of DeliriUM in ICU (PODIUM) - BMJ Open — bmjopen.bmj.com — https://bmjopen.bmj.com/content/bmjopen/13/7/e072095.full.pdf
2. Adverse Outcomes in Critically Ill Patients with Delirium | NEJM Clinician — clinician.nejm.org — https://clinician.nejm.org/adverse-outcomes-critically-ill-patients-delirium-nejm-jw.NA38131
3. Long-Term Cognitive Impairment after Critical Illness — www.nejm.org — https://www.nejm.org/doi/full/10.1056/NEJMoa1301372
4. Effectiveness of non-pharmacological intervention protocol for prevention of postoperative delirium in the surgical intensive care unit | Scientific Reports — www.nature.com — https://www.nature.com/articles/s41598-025-86768-4
5. SCCM Guidelines : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/collections/8
6. A Focused Update to the Clinical Practice... : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/10.1097/ccm.0000000000006574~a-focused-update-to-the-clinical-practice-guidelines-for-the
7. Society of Critical Care Medicine Guidelines on... : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/2026/05000/society_of_critical_care_medicine_guidelines_on.3.aspx
8. Society of Critical Care Medicine Guidelines for... : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/2026/03000/society_of_critical_care_medicine_guidelines_for.20.aspx
9. Comparative effectiveness of non-pharmacological interventions for preventing delirium in critically ill adults: A systematic review and network meta-analysis — www.sciencedirect.com — https://www.sciencedirect.com/science/article/abs/pii/S0020748922000682
10. Haloperidol - an overview | ScienceDirect Topics — www.sciencedirect.com — https://www.sciencedirect.com/topics/medicine-and-dentistry/haloperidol
11. 10 key issues for prevention, monitoring and non-pharmacological treatment of delirium in critically ill patients - ScienceDirect — www.sciencedirect.com — https://www.sciencedirect.com/science/article/abs/pii/S2529984025000151
12. Impact of consultation liaison services on postoperative psychotropic drug use — www.sciencedirect.com — https://www.sciencedirect.com/science/article/pii/S002239992500087X
13. [PDF] Protocol for - ClinicalTrials.gov — cdn.clinicaltrials.gov — https://cdn.clinicaltrials.gov/large-docs/76/NCT07649876/Prot_000.pdf
14. Prevention and management of delirium in critically ill adult patients in the intensive care unit: a review based on the 2018 PADIS guidelines - PMC — www.ncbi.nlm.nih.gov — https://www.ncbi.nlm.nih.gov/pmc/articles/6786674
15. DECREASING DELIRIUM THROUGH MUSIC IN ... — cdn.clinicaltrials.gov — https://cdn.clinicaltrials.gov/large-docs/34/NCT04182334/Prot_SAP_000.pdf
16. Prevention and management of delirium in critically ill adult patients in the intensive care unit: a review based on the 2018 PADIS guidelines - PubMed — www.ncbi.nlm.nih.gov — https://www.ncbi.nlm.nih.gov/pubmed/31723916
17. [PDF] June 2023 - CDC — wwwnc.cdc.gov — https://wwwnc.cdc.gov/eid/content/29/6/pdfs/v29-n6.pdf
18. Impact of the 2018 Society of Critical Care Medicine Pain,... : Critical Care Explorations — journals.lww.com — https://journals.lww.com/ccejournal/fulltext/2021/10000/impact_of_the_2018_society_of_critical_care.24.aspx
19. Melatonin Use in the ICU: A Systematic Review and... : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/2025/09000/melatonin_use_in_the_icu__a_systematic_review_and.1.aspx
20. Poststroke delirium incidence and outcomes : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/abstract/10.1097/ccm.0b013e318232da12~poststroke-delirium-incidence-and-outcomes-validation-of-the
21. Clinical Practice Guidelines for the Prevention... : Critical Care ... — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/10.1097/ccm.0000000000003299~clinical-practice-guidelines-for-the-prevention-and
22. Evaluation of delirium in critically ill... : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/2001/07000/evaluation_of_delirium_in_critically_ill_patients_.12.aspx
23. Abstracts From the 51st Annual Meeting of the... : Journal of Neurosurgical Anesthesiology — journals.lww.com — https://journals.lww.com/jnsa/fulltext/2023/10000/abstracts_from_the_51st_annual_meeting_of_the.15.aspx
24. A Novel Computerized Test for Detecting and... : Critical Care Medicine — journals.lww.com — https://journals.lww.com/ccmjournal/fulltext/10.1097/ccm.0000000000002477~a-novel-computerized-test-for-detecting-and-monitoring

## Editorial note

Prepared from cited clinical literature using Astra's research workflow. Verify recommendations against current guidance and patient-specific factors.