Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with the interim process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.
Published evidence
Three studies were identified by the literature search and are summarised in this briefing. They include a total of 157 people admitted to intensive care, 137 of whom had monitoring with CerebAir.
The clinical evidence and its strengths and limitations is summarised in the overall assessment of the evidence.
Overall assessment of the evidence
The evidence base is limited and mainly comes from single-centre observational studies that involve a relatively small number of people. The studies mainly assessed the feasibility of using the technology in an acute setting and its diagnostic accuracy compared with routine electroencephalogram (EEG) only. None of the studies reported about changes in clinical management because of CerebAir monitoring. Two of the 3 studies were prospective in design. None of the studies were done in the UK.
Overall, the evidence base suggests that CerebAir is faster to position than standard EEG techniques and is likely to be feasible in an acute setting. However, clinical expert advice was that the reduction in the time to apply CerebAir as shown in Caricato et al. (2020) is unlikely to be clinically significant. The technology appears to be reliable in detecting abnormal EEG patterns, with Egawa et al. (2020) reporting a sensitivity and specificity for detecting abnormal EEG patterns of 0.974 and 0.909, respectively. Caricato et al. (2020) reported that more people having monitoring with CerebAir experienced skin redness. No adverse reactions were observed by Egawa et al. (2020) and the remaining study did not report this outcome.
There is limited evidence on CerebAir for continuous EEG monitoring. The evidence base would benefit from further prospective evidence including a larger cohort of patients, ideally done in the UK. Also, future studies evaluating the impact of using the technology on patient outcomes and changes in clinical management would be useful.
Caricato et al. (2020)
Intervention and comparator
CerebAir compared with conventional simplified EEG (8‑electrodes‑EEG positioned by an EEG technician).
Key outcomes
The time needed to apply electrodes was shorter with CerebAir than with the control (6.2 minutes compared with 10.4 minutes, p<0.001). Length of monitoring was shorter with CerebAir (57 hours compared with 75 hours, p<0.001) but was longer than 24 hours in 43% of people (n=13). CerebAir needed more interventions per person to correct artifacts and get good quality EEG recordings (1.7 compared with 0.5, p<0.001). A total of 35 interventions (4 electrode replacements and 31 gel or paste applications) were needed with CerebAir compared with 11 interventions (3 electrode replacements and 8 gel or paste applications) with the control. EEG abnormalities were detected in 14 people in the CerebAir (7 people with epilepsy and 7 without) and control group (5 people with epilepsy and 9 without). EEG recordings led to antiseizure medicines in 10 people in the CerebAir group compared with 7 cases in the control group. Seventeen people in the CerebAir group experienced skin redness because of pressure lesions. Lesions appeared after a mean time of 15 hours and resolved without further intervention. In 4 people, EEG monitoring was stopped because of more severe pressure lesions. This happened after a mean time of 52 hours.
Strengths and limitations
Study groups were similar in terms of age, gender, diagnosis, length of intensive care stay and likelihood of recovery.
The sample size was small, and the 2 groups were also treated on different intensive care wards (neuro and general). The study was not powered to show differences in detected EEG abnormalities between the 2 groups. No conclusions can be drawn about the accuracy of the technology for seizure diagnosis compared with conventional EEG. The study was done in Italy and may limit the generalisability to the NHS. One of the authors is an associate editor of the journal. The EEG recordings were reviewed by an expert neurologist and not at the bedside by an intensivist.
Egawa et al. (2020)
Intervention and comparator
AE‑120A EEG Headset (CerebAir) compared with conventional continuous EEG monitoring, both with a video camera monitoring.
Key outcomes
Of the 65 people who were monitored with CerebAir, 50 (76.9%) were included in the final analyses (median age of 72 years; 66% were male). The sensitivity and specificity of CerebAir for detecting abnormal EEG patterns were 0.974 (95% confidence interval [CI] 0.865 to 0.999) and 0.909 (95% CI 0.587 to 0.998), respectively. The sensitivity and specificity of CerebAir for detecting periodic discharges were 0.824 (95% CI 0.566 to 0.926) and 0.970 (95% CI 0.842 to 0.999), respectively. Thirteen people (26%) were diagnosed with nonconvulsive status epilepticus (NCSE) using CerebAir, and the technology could detect NCSE with a sensitivity and specificity of 0.706 (95% CI 0.440 to 0.897) and 0.970 (95% CI 0.842 to 0.999), respectively. The median time to start monitoring with CerebAir was 57 minutes (ranging from 5 to 142 minutes). No adverse reactions were seen.
Strengths and limitations
EEG recordings were interpreted by 1 neurointensivist as well as a board-certified neurophysiologist.
The study included a relatively small number of people. It was done prospectively at a single centre but retrospectively analysed, which may have introduced potential selection bias. Interventions were not used simultaneously, meaning differences could have been caused by the time interval between recordings and clinical interventions used. The median monitoring time with CerebAir was only 134.5 minutes. A longer monitoring time may have been needed to detect NCSE. The study was done in Japan and may limit the generalisability to the NHS. There is limited information on people with acute brain injury which may be of interest to intensivists working in the NHS.
Meyer et al. (2021)
Key outcomes
There were 47 people included in the final per-protocol analysis. Five people did not have routine EEG because of technical issues or medical conditions. The agreement between CerebAir and routine EEG for EEG background activity, epileptiform discharges and seizure activity was 53% (24 of 45 people; p=0.126), 68% (32 people; p=0.162) and 98% (46 people; p value not stated), respectively. Compared with routine EEG, CerebAir detected the same or additional intensive care-relevant EEG patterns in 89% of people.
Strengths and limitations
Patients' conditions were typical of those in neurointensive care. CerebAir recordings were analysed by physicians blinded to routine EEG results. It was unclear from the study whether physicians were specialist neurophysiologists or intensivists.
Analysis was done per protocol because, out of the 52 people in the study, only 47 people had both interventions. EEG activity may have been affected by adjustments of antiseizure and sedative medicines as well as the time between recordings with the 2 interventions. It was not documented whether using CerebAir led to treatment changes, so the clinical consequence of using the technology is not known. The study was done in Germany and may limit the generalisability to the NHS.