Continuous EEG & ICU Monitoring
Continuous EEG (cEEG) monitoring in the intensive care unit has become an indispensable tool for the detection and management of seizures in critically ill patients. Seizures are frequently encountered in ICU populations, with reported prevalence rates of 10–35% depending on the underlying etiology and the criteria used for seizure definition. The majority of these seizures are nonconvulsive and would be undetectable without EEG monitoring. The American Clinical Neurophysiology Society (ACNS) has developed a standardized terminology system for describing ICU EEG patterns, which has improved communication among clinicians and facilitated research into the clinical significance of periodic and rhythmic patterns that exist on the ictal-interictal continuum. Understanding these patterns, their clinical implications, and the evolving role of quantitative EEG (qEEG) trending tools is essential for neurologists and intensivists caring for critically ill patients.
Bottom Line
- Nonconvulsive seizures are common in critically ill patients (10–35% of monitored ICU patients) and are undetectable without EEG monitoring
- ACNS standardized terminology provides a systematic framework for describing periodic patterns (LPDs, GPDs, BIPDs), rhythmic delta activity (LRDA, GRDA), and seizures in the ICU setting
- The ictal-interictal continuum describes patterns that do not meet seizure criteria but carry a risk of neuronal injury and may warrant treatment; they often correlate with intracranial seizures and increased metabolic demands
- The 2HELPS2B score predicts seizure risk and guides monitoring duration; a score ≥2 supports 24-hour cEEG recording
- Quantitative EEG (qEEG) trending tools compress hours of data into a single screen, enabling faster screening for seizures and trending of seizure burden; raw EEG review remains the gold standard when a seizure is suspected
- Seizure burden in the ICU is associated with worse functional outcomes; aggressive treatment of seizures and highly epileptiform patterns is generally recommended, though optimal treatment strategies for the ictal-interictal continuum remain debated
Indications for Continuous EEG Monitoring
The decision to initiate cEEG monitoring should be based on clinical risk factors for seizures in the ICU setting. The following table outlines the major indications and clinical scenarios.
| Indication | Clinical Scenario | Expected Utility |
|---|---|---|
| Unexplained altered mental status | Comatose or encephalopathic patient without clear explanation; mental status disproportionate to known insult | Detect nonconvulsive seizures (NCS) or nonconvulsive status epilepticus (NCSE) as contributing factor |
| Status epilepticus | After convulsive SE is treated; clinical seizure activity resolves but mental status does not return to baseline | Monitor for ongoing electrographic seizures (present in up to 48% after convulsive SE is treated); titrate treatment |
| Acute brain injury | Subarachnoid hemorrhage, intracerebral hemorrhage, traumatic brain injury, ischemic stroke, CNS infection | Detect secondary seizures; NCS rate is 10–30% in SAH and ICH; monitor for delayed cerebral ischemia (in SAH) |
| Post-cardiac arrest | Comatose survivors of cardiac arrest during targeted temperature management | Detect subclinical seizures or myoclonus; background reactivity and continuity are prognostic markers |
| Acute seizure flurries | Patient with multiple clinical seizures in the ED or ICU; concern for ongoing subclinical seizures | Quantify seizure burden; assess response to treatment; detect progression to NCSE |
| Pharmacologic coma monitoring | Refractory SE treated with continuous anesthetic infusions (midazolam, propofol, pentobarbital) | Titrate infusion to burst-suppression or seizure-suppression target; guide weaning attempts |
ACNS Standardized Critical Care EEG Terminology
The ACNS standardized terminology (2021 revision) provides a systematic language for describing the electrographic patterns commonly encountered in critically ill patients. This system classifies patterns by their spatial distribution (lateralized, generalized, bilateral independent), morphology, frequency, and modifiers.
Classification Framework
| Main Term | Abbreviation | Definition | Clinical Association |
|---|---|---|---|
| Lateralized periodic discharges | LPDs | Periodic sharply contoured discharges confined to one hemisphere; frequency typically 0.5–3 Hz | Acute structural lesion (stroke, tumor, encephalitis); high seizure risk (~60%); formerly "PLEDs" |
| Generalized periodic discharges | GPDs | Periodic discharges with a generalized (bilateral, bisynchronous) distribution | Metabolic encephalopathy (when triphasic morphology), hypoxic-ischemic injury, prion disease (CJD), drug toxicity, NCSE |
| Bilateral independent periodic discharges | BIPDs | Periodic discharges occurring independently in each hemisphere; different frequencies or morphologies | Severe diffuse brain injury; CNS infection (especially HSV encephalitis — classic association); high seizure risk and poor prognosis |
| Lateralized rhythmic delta activity | LRDA | Rhythmic delta activity confined to one hemisphere; frequency <4 Hz | Focal structural or epileptic pathology; seizure risk intermediate between LPDs and no epileptiform findings |
| Generalized rhythmic delta activity | GRDA | Rhythmic delta activity with generalized distribution; typically frontally predominant | Encephalopathy (metabolic, toxic, infectious); lower seizure risk than GPDs; includes pattern previously called FIRDA |
| Electrographic seizure | Sz | Epileptiform discharges averaging >2.5 Hz for ≥10 seconds; OR any pattern with definite evolution lasting ≥10 seconds; OR pattern with clinical correlate time-locked to EEG | Requires treatment; may be convulsive or nonconvulsive; evolution in frequency, morphology, or distribution is the hallmark |
Important Modifiers
- Frequency: Reported in Hz (e.g., 1 Hz LPDs, 2 Hz GPDs); frequency ≥2.5 Hz meets seizure criteria when periodic/spike-wave
- "Plus" modifier (+): Superimposed fast activity or rhythmic activity on periodic discharges (e.g., LPDs+, GPDs+); the "plus" modifier indicates higher ictal potential and increased seizure risk
- Triphasic morphology: Applied to GPDs with smooth, blunted waveforms, three phases with progressively longer durations, and anterior-to-posterior time lag; classically associated with metabolic encephalopathy but does not exclude an epileptogenic process
- Prevalence: Abundant (≥50% of recording), frequent (10–49%), occasional (1–9%), rare (<1%)
- Stimulus-induced/terminated: Patterns that appear or resolve with stimulation (stimulus-induced rhythmic, periodic, or ictal discharges — SIRPIDs)
The Ictal-Interictal Continuum
The ictal-interictal continuum (IIC) is a concept that acknowledges the spectrum between clearly interictal patterns and definite electrographic seizures. Many periodic and rhythmic patterns in the ICU do not meet formal ACNS criteria for seizure but carry a meaningful risk of neuronal injury and are associated with worse outcomes.
Patterns on the Ictal-Interictal Continuum
- Periodic discharges with frequency between 1 Hz and 2.5 Hz (do not meet ≥2.5 Hz seizure threshold)
- LPDs, GPDs, and BIPDs — especially with the "plus" modifier (superimposed fast or rhythmic activity)
- LRDA, particularly when sharply contoured or with admixed spikes
- Stimulus-induced patterns (SIRPIDs)
- These patterns have been shown to correlate with electrographic seizures on intracranial recording and with increased cerebral metabolic demands on microdialysis studies
- Treatment decisions for IIC patterns depend on multiple variables: clinical context, underlying diagnosis, response to antiseizure medication trials, and neuroimaging findings
- A pragmatic approach: consider treatment if the pattern resolves with an ASM trial and the patient's clinical examination improves, or if there is evidence of ongoing metabolic crisis on multimodality monitoring
Nonconvulsive Seizures and NCSE
Nonconvulsive seizures (NCS) and nonconvulsive status epilepticus (NCSE) are electrographic seizures that occur without prominent motor manifestations. They are a major indication for cEEG monitoring because they are essentially undetectable at the bedside without EEG.
Prevalence and Risk Factors
- NCS are detected in 10–35% of critically ill patients undergoing cEEG monitoring
- Higher rates are reported in specific populations: subarachnoid hemorrhage (10–30%), intracerebral hemorrhage (15–25%), traumatic brain injury (15–25%), post-cardiac arrest (10–35%), and after convulsive status epilepticus (up to 48%)
- Risk factors include: acute brain injury, prior seizure history, younger age, coma, and the presence of periodic discharges on early EEG
NCSE Diagnostic Criteria
- The Salzburg Consensus Criteria for NCSE require electrographic seizure activity lasting ≥10 minutes, or for ≥20% of any 60-minute recording period
- Modified criteria also allow diagnosis if the clinical picture is consistent with NCSE and the EEG pattern meets seizure criteria for any duration with an unequivocal clinical correlate
- Subtle clinical signs may include eye deviation, nystagmus, facial twitching, or finger automatisms
Clinical Significance
- NCS are independently associated with increased morbidity and mortality in critically ill patients
- Duration of electrographic seizures correlates with worse outcomes; early detection and treatment is therefore essential
- NCSE can cause secondary brain injury through excitotoxicity, metabolic exhaustion, and blood-brain barrier disruption, even in the absence of motor convulsions
The 2HELPS2B Score
The 2HELPS2B score is a validated clinical prediction tool that helps clinicians estimate the risk of electrographic seizures in ICU patients, guiding decisions about cEEG monitoring duration.
| Feature | Points |
|---|---|
| Brief potentially ictal rhythmic discharges (B(I)RDs) | 2 |
| Lateralized periodic discharges (LPDs), lateralized rhythmic delta activity (LRDA), or bilateral independent periodic discharges (BIPDs) | 2 |
| Prior seizure (history of epilepsy or clinical seizure during this admission) | 1 |
| Sporadic epileptiform discharges | 1 |
| Frequency >2 Hz of any periodic/rhythmic pattern | 1 |
| "Plus" features (superimposed fast or rhythmic activity on periodic discharges) | 1 |
Interpreting the 2HELPS2B Score
- Score 0: Seizure risk approximately 5% — brief monitoring may suffice; consider discontinuation if no epileptiform features after initial screening
- Score 1: Seizure risk approximately 12% — monitoring for at least 12–24 hours is reasonable
- Score ≥2: Seizure risk approximately 25–95% (increases with higher scores) — 24-hour or longer monitoring is recommended
- In the absence of interictal findings indicating an increased risk of seizure, 24 to 48 hours of EEG is often acquired, depending on the clinical status of the patient
- The score has been validated in multiple cohorts and is increasingly used to standardize monitoring decisions
Quantitative EEG (qEEG) Trending Tools
Quantitative EEG refers to computerized mathematical processing of raw EEG data that generates compressed displays for rapid screening of prolonged recordings. These tools do not replace raw EEG review but serve as a supplement to facilitate efficient monitoring.
Common qEEG Displays
| qEEG Tool | What It Measures | Clinical Utility |
|---|---|---|
| Color density spectral array (CDSA) | Power spectral density over time, displayed as color (typically warm colors = high power); frequency on y-axis, time on x-axis | Seizures appear as bright streaks or arches of increased power; allows rapid scanning of hours of data on a single screen |
| Amplitude-integrated EEG (aEEG) | Semi-logarithmic display of peak-to-peak amplitude over time | Widely used in neonatal ICU; identifies seizures (acute rises in amplitude), burst-suppression, and background discontinuity |
| Asymmetry index | Hemispheric power ratio over time | Detects lateralized changes suggesting focal seizures, new structural lesions, or delayed cerebral ischemia (in SAH) |
| Alpha-delta ratio | Ratio of alpha power to delta power per hemisphere over time | Decreasing ratio indicates encephalopathy or cerebral ischemia; used for vasospasm detection in SAH |
| Rhythmicity spectrogram | Quantifies periodic and rhythmic patterns by frequency and time | Highlights seizures and periodic patterns; can track response to treatment over hours |
- qEEG can reduce review time for continuous EEG studies, especially those with frequent seizures
- The compression of multiple hours of data to a single screen makes it easier to see overarching trends in seizure frequency and response to treatment
- Important caveat: It remains essential to review the raw EEG data whenever a seizure is suspected on qEEG — artifacts can mimic seizures on compressed displays
Duration of Monitoring
The optimal duration of cEEG monitoring depends on the clinical indication, the initial EEG findings, and the patient's evolving clinical course.
- Minimum screening: If no epileptiform features or seizures are detected in the first 30–60 minutes, the yield of continued monitoring decreases but does not reach zero
- Standard duration: Most guidelines recommend at least 24 hours of monitoring for patients at moderate-to-high risk for NCS
- Extended monitoring (>48 hours): Indicated for patients with persistent altered mental status, ongoing periodic/rhythmic patterns, breakthrough seizures despite treatment, or pharmacologic coma for refractory SE
- Seizure risk decreases over time: Among patients who will have seizures detected by cEEG, the majority (approximately 80–90%) are detected within the first 24 hours; however, delayed seizures can occur, particularly in SAH
- The 2HELPS2B score helps guide duration decisions — patients with higher scores warrant longer monitoring
Seizure Burden and Outcomes
The cumulative burden of electrographic seizures in the ICU is an emerging concept that has important implications for prognosis and treatment intensity.
- Definition: Seizure burden is typically expressed as the percentage of time spent in seizure (e.g., minutes of seizure activity per hour of monitoring)
- Outcome data: Multiple studies have demonstrated that higher seizure burden is independently associated with worse functional outcomes, including increased mortality and disability
- In subarachnoid hemorrhage, epileptiform abnormality burden has been associated with worse neurologic outcomes and changes in antiepileptic drug management
- The relationship between periodic pattern burden (not meeting seizure criteria) and outcomes suggests that even sub-seizure patterns may contribute to secondary brain injury
- Treatment target: Most epileptologists aim to reduce seizure burden to <10% of the monitoring period; complete elimination of seizures is the ideal goal when achievable without excessive sedation
Challenges and Pitfalls in ICU EEG Monitoring
- Artifact burden: The ICU environment generates extensive artifact (ventilator, dialysis machine, chest percussion, electrical equipment, patient movement); artifact can mimic seizures and obscure true epileptiform activity
- Rapidly applied/limited montage EEG: Several rapid-application EEG systems with limited electrodes have been developed for urgent ICU use; however, limited montages can miss focal activity distant from recording electrodes — the standard 10–20 montage remains the gold standard
- Medication effects: Sedative infusions (propofol, midazolam) suppress electrographic seizures and may mask the true seizure risk; burst-suppression from anesthetics is a treatment target in refractory SE but obscures background interpretation
- Triphasic waves vs. NCSE: GPDs with triphasic morphology can be extremely difficult to distinguish from NCSE; trial of benzodiazepine with assessment for both EEG and clinical improvement may help, but response to benzodiazepine alone does not confirm NCSE (metabolic patterns may also improve)
- Stimulus-induced patterns: SIRPIDs may be provoked by routine nursing care (suctioning, repositioning); their significance is debated, and they should not be confused with spontaneous seizures
Practical Workflow for ICU cEEG
- Initiation: Apply standard 10–20 electrodes using collodion or paste; ensure electrode impedances are <5 kΩ; document patient's baseline neurologic examination
- First hour: Technologist or neurophysiologist reviews initial recording for epileptiform activity, periodic patterns, and background assessment; determine monitoring tier using 2HELPS2B or institutional protocol
- Ongoing review: Continuous review by trained staff or remote monitoring service; real-time seizure detection algorithms with technologist verification; qEEG trending reviewed every 2–4 hours
- Reporting: Initial report within the first 1–2 hours; daily summary reports describing seizure frequency, seizure burden, periodic pattern prevalence, background assessment, and clinical correlations
- Communication: Immediate notification to clinical team for new electrographic seizures, NCSE, or significant changes in periodic pattern frequency
- Discontinuation: Consider when seizures are controlled for ≥24 hours, periodic patterns resolve, pharmacologic coma is weaned without seizure recurrence, or clinical goals change to comfort care
References
- Hirsch LJ, Fong MWK, Leitinger M, et al. American Clinical Neurophysiology Society's standardized critical care EEG terminology: 2021 version. J Clin Neurophysiol. 2021;38(1):1–29.
- Claassen J, Mayer SA, Kowalski RG, Emerson RG, Hirsch LJ. Detection of electrographic seizures with continuous EEG monitoring in critically ill patients. Neurology. 2004;62(10):1743–1748.
- Struck AF, Tabaeizadeh M, Schmitt SE, et al. Assessment of the validity of the 2HELPS2B score for inpatient seizure risk prediction. JAMA Neurol. 2020;77(4):500.
- Rubinos C, Reynolds AS, Claassen J. The ictal-interictal continuum: to treat or not to treat (and how)? Neurocrit Care. 2018;29(1):3–8.
- Claassen J, Perotte A, Albers D, et al. Nonconvulsive seizures after subarachnoid hemorrhage: multimodal detection and outcomes. Ann Neurol. 2013;74(1):53–64.
- Holla SK, Krishnamurthy PV, Subramaniam T, Dhakar MB, Struck AF. Electrographic seizures in the critically ill. Neurol Clin. 2022;40(4):907–925.
- Zafar SF, Postma EN, Biswal S, et al. Effect of epileptiform abnormality burden on neurologic outcome and antiepileptic drug management after subarachnoid hemorrhage. Clin Neurophysiol. 2018;129(11):2219–2227.
- Moura LMVR, Shafi MM, Ng M, et al. Spectrogram screening of adult EEGs is sensitive and efficient. Neurology. 2014;83(1):56–64.
- Krumholz A. Nonepileptic seizures: diagnosis and management. Neurology. 1999;53:s76–s78.
- Arroyo S, Lesser RP, Fisher RS, et al. Clinical and electroencephalographic evidence for sites of origin of seizures with diffuse electrodecremental pattern. Epilepsia. 1994;35(5):974–987.
- Fernández-Torre JL, Kaplan PW. Atypical or typical triphasic waves — is there a difference? J Clin Neurophysiol. 2021;38(5):384–398.
- Leitinger M, Trinka E, Gardella E, et al. Diagnostic accuracy of the Salzburg EEG criteria for non-convulsive status epilepticus. Lancet Neurol. 2016;15(13):1054–1062.
- Vespa PM, Nuwer MR, Nenov V, et al. Increased incidence and impact of nonconvulsive and convulsive seizures after traumatic brain injury as detected by continuous electroencephalographic monitoring. J Neurosurg. 1999;91(5):750–760.
- Herman ST, Abend NS, Bleck TP, et al. Consensus statement on continuous EEG in critically ill adults and children. Part I: indications. J Clin Neurophysiol. 2015;32(2):87–95.
- Tsuchida TN, Wusthoff CJ, Shellhaas RA, et al. American Clinical Neurophysiology Society standardized EEG terminology and categorization for the description of continuous EEG monitoring in neonates. J Clin Neurophysiol. 2013;30(2):161–173.
