Table 1.
Key study characteristics of studies measuring EEG prior to delirium.
| Study | Sample characteristics | Study quality | Delirium outcome and measure | EEG characteristics | Main findings relative to the presence of delirium |
|---|---|---|---|---|---|
| Andresen et al. (2014) | Participants n = 69 mechanically ventilated ICU patients Delirium n = 42, non-delirious matched control n = 27 Age = 57 (6.4) Sex (F) = 48% Country: USA |
5/5 | Outcome: Incidence Measure: CAM-ICU |
4-channel BIS Unspecified montage Burst suppression Percent of previous 63-second epoch that is isoelectric |
↑Time in burst suppression |
| Azabou et al. (2015) | Participants n = 110 (septic shock n = 45, severe sepsis n = 37, systemic inflammatory response syndrome n = 28) Delirium n = 22, non-delirious matched control n = 42, sedated n = 46 Age = 63.8 (18.1) Sex (F) = 29% Country: USA |
5/5 | Outcome: Presence Measure: CAM-ICU |
11-channel EEG 10/20 montage 20-minute recording Synek & Young EEG classification scale Delta power (did not state if relative or absolute) |
↑Electrographic seizures ↑ Slowing |
| Fritz et al. (2016) | Participants n = 619 receiving general anaesthesia Delirium n = 162, non-delirious matched control n = 457 Age = 62 (14) Sex (F) = 36% Country: USA |
5/5 | Outcome: Presence Measure: CAM-ICU |
1-channel BIS Unspecified montage Measured continuously throughout surgery Burst suppression Percent of preceding 63 seconds for which EEG was electrically suppressed |
BIS values < 20 ↑ Time in burst suppression |
| Fritz et al. (2018) | Participants n = 618 undergoing elective surgery Delirium n = 162, non-delirious matched control n = 456 Age = 62 (range 18-92) Sex (F) = 36% Country: USA |
5/5 | Outcome: Presence Measure: CAM-ICU |
1-channel BIS Monitored continuously throughout surgery Burst suppression Percentage of preceding 63 seconds for which the EEG amplitude was < 5 microvolts, captured once per minute |
Presence of burst suppression |
| Hesse et al. (2018) | Participants n = 626 receiving general anaesthesia Delirium n = 125, non-delirious matched control n = 501 Age = 56.9 (6.9) Sex (F) = 39% Country: USA |
4/5 | Outcome: Presence Measure: CAM-ICU |
BIS (electrode number unspecified) Recorded throughout surgery Burst suppression 10 second episodes (episode defined as burst suppression if one bursting period could be detected between two suppression episodes) |
Presence of burst suppression |
| Koch et al. (2018) | Participants n = 62 children undergoing planned surgery Delirium n = 27, non-delirious matched control n = 35 Age = 3.8 (2.7) Sex (F) = 40.3% Country: Germany |
5/5 | Outcome: Presence Measure: Paediatric assessment of emergence delirium score |
4-channel EEG Recorded from before start of anaesthesia until end of anaesthesia Epoch and amount of data not specified Epileptiform discharges, DSP, PSR, PED and SSP patterns |
↑Epileptiform discharges (rhythmic polyspikes, periodic epileptiform discharges and delta with spikes) |
| Martin et al. (2014) | Participants n = 12 children during emergence of anaesthesia Delirium n = 5, non-delirious normal control n = 7 Age = control median 6.6 (0.8), delirium median 5.7 (0.9) Sex (F) = 25% Country: Australia |
3/5 | Outcome: Presence Measure: Authors own classification |
64-channel EEG 10/5 montage Recorded before discontinuation of anaesthetic Five-minute time bins Functional connectivity (Global efficiency and global coherence) |
↑Frontal region global efficiency Diffused mixed alpha and beta connectivity |
| Muhlhofer et al. (2017) | Participants n = 41 undergoing general anaesthesia Delirium n = 7, non-delirious control n = 34 Age = 62.3 (8.6) Sex (F) = 51.2% Country: USA |
5/5 | Outcome: Presence Measure: CAM |
4-channel BIS 10/20 system Recorded during surgery 30-second epochs Burst suppression indicated by percentage of complete EEG suppression during one minute |
↑Time in burst suppression |
| Numan et al. (2019) | Participants n = 159 undergoing surgery Delirium n = 29, possible delirium n = 26, non-delirious matched control n = 104 Age = 76.9 (6.2) Sex (F) = 33.3% Country: Netherlands |
5/5 | Outcome: Presence Measure: DRS-R-98, CAM-ICU |
3-channel EEG 10/20 montage 5-minute recording 60 seconds artefact free data selected Relative delta power |
n.s. Relative delta power |
| Radtke et al. (2013) | Participants n = 1155 for general anaesthesia (BIS guided n = 575, BIS blinded n = 580) Delirium n = 219, non-delirious control n = 936 Age = 69.9 (6.4) Sex (F) = 46.1% Country: Germany |
5/5 | Outcome: Presence Measure: DSM-IV |
2-channel BIS Unspecified montage Data recorded at minimum intervals of 1 minute BIS values |
BIS values < 20 |
| Schramm et al. (2017) | Participants n = 50 ICU patients Delirium n = 19, non-delirious matched control n = 31 Age = 63 (19) Sex (F) = 30% Country: Germany |
4/5 | Outcome: Incidence Measure: CAM-ICU |
10-channel EEG 10/20 montage 24-h recording Epochs not specified Classified by severity based on predominant waveform grave I t o V |
n.s. Background activity, burst suppression or suppressed background activity |
| Sieber et al. (2010) | Patients n = 114 undergoing hip fracture repair (deep sedation n = 57, light sedation n = 57) Delirium n = 34, non-delirious matched control = 80 Age = 81.5 (7.1) Sex (F) = 72.8% Country: USA |
5/5 | Outcome: Presence Measure: CAM |
4-channel BIS Unspecified montage Amount of data unspecified BIS values |
n.s. BIS values |
| Soehle et al. (2015) | Participants n = 81 undergoing cardiac surgery Delirium n = 26, non-delirious matched control n = 55 Age = 72.9 (6.2) Sex (F) = 29.6% Country: Germany |
4/5 | Outcome: Incidence Measure: CAM-ICU flowchart |
2-channel BIS Unspecified montage Sampled in 5-second intervals Burst suppression ratio defined as percentage of epochs in previous 63 seconds that are suppressed BIS values, burst suppression, EEG asymmetry (In total power within 0–30 Hz) |
↑Time in burst suppression n.s. Hemispheric asymmetry n.s. BIS values |
Note. Age = mean (standard deviation) unless stated otherwise.
↑ indicates statistically significant increase; ↓indicates statistically significant decrease; n.s. indicates non-significant relationship between delirium and EEG measure.
CAM-ICU = Confusion Assessment Method for Intensive Care Unit; BIS = Bispectral Index Monitoring; EEG = Electroencephalogram; DRS-R-98 = Delirium Rating Scale-Revised-98; DSV-IV = Diagnostic and Statistical Manual of Mental Disorders 4th Edition; DSP = delta with spikes; PSR = rhythmic polyspikes; periodic epileptiform discharges; SSP = suppression with spikes; ICU = Intensive Care Unit; USA = United States of America.