Can You See Delirium on a Full Montage EEG?
Yes, delirium produces characteristic EEG abnormalities that can be identified on full-montage EEG, making it a valuable objective tool for supporting the clinical diagnosis, though EEG findings alone cannot definitively diagnose delirium without clinical correlation.
What EEG Shows in Delirium
The EEG during delirium demonstrates several characteristic features 1:
- Diffuse slowing of background activity - the most consistent finding
- Increased spectral variability
- Periodic discharges including triphasic waves
- Polymorphic delta activity
- In some cases: electrographic seizures or periodic epileptiform discharges
A critical caveat: A normal EEG essentially excludes delirium 2, 3. This makes EEG particularly useful as a rule-out tool when the clinical picture is unclear.
Specific EEG Patterns That Identify Delirium
Recent research has identified quantifiable EEG metrics with excellent diagnostic performance 4:
- Delta variability (increased)
- High-beta variability (increased)
- Relative theta power (increased)
- Relative alpha power (decreased)
- Reduced beta power - a novel specific finding 2
These four metrics combined achieved an area under the curve of 0.94 for distinguishing delirious from non-delirious ICU patients 4. The combination of spectral power at F3-P4 at 2 Hz and C3-O1 at 19 Hz provided 100% sensitivity and 99% specificity in identifying delirium among normal controls 2.
Important Clinical Considerations
When EEG is Most Useful
EEG serves multiple critical functions in delirium assessment 1:
- Ruling out non-convulsive status epilepticus (NCSE) - Many patients with NCSE present with delirium, and this must be considered in high-risk patients
- Supporting uncertain clinical diagnoses - When behavioral assessment tools are equivocal
- Monitoring patients who cannot be assessed clinically - Those with severely reduced arousal who are non-comatose but non-communicative
Key Pitfalls to Avoid
Stroke patients require special attention 5. In patients with right hemispheric stroke, single-channel EEG systems have higher false-positive rates. Slowing in delta or theta frequencies over the stroke site can mimic delirium patterns. Use bihemispheric recordings in stroke patients rather than unilateral recordings to avoid misdiagnosis.
Different etiological subtypes of delirium (toxic-metabolic, infectious, postoperative) are assumed to share similar EEG features, though this has not been extensively studied 1.
Practical Application
Full Montage vs. Simplified EEG
While full-montage EEG provides comprehensive data, simplified point-of-care EEG systems with limited electrodes are increasingly validated 3, 6. These portable systems:
- Allow rapid bedside deployment
- Can be used by non-specialist staff
- Show acceptable diagnostic accuracy (though specificity around 71% with some systems 5)
- Are more practical for repeated assessments in general wards
However, for definitive assessment, particularly in complex cases or stroke patients, full-montage EEG remains preferable 5.
Integration with Clinical Assessment
EEG should complement, not replace, clinical delirium assessment tools like CAM-ICU 7. CAM scores correlate strongly with EEG abnormalities (r = 0.639) and predict EEG changes 7. The combination provides:
- Objective neurophysiological data (EEG)
- Standardized behavioral assessment (CAM)
- Enhanced diagnostic confidence
Age Considerations
Patients aged 75 and older show the highest incidence of EEG abnormalities in delirium (88.2%) 7, making EEG particularly valuable in elderly populations where clinical assessment may be complicated by baseline cognitive impairment.
Bottom Line for Clinical Practice
Order a full-montage EEG when:
- Clinical delirium assessment is uncertain or fluctuating
- You need to rule out NCSE in high-risk patients
- The patient has severely reduced arousal but is not comatose
- You need objective documentation to guide treatment decisions
Interpret with caution when:
- The patient has acute stroke (use bihemispheric analysis)
- Sedating medications are on board (may confound findings)
- Baseline cognitive impairment exists (establish comparison if possible)
Remember: EEG abnormalities in delirium have prognostic value - higher EEG-based delirium scores correlate with severity and predict adverse outcomes including reduced survival 6.