Continuous EEG Randomized Trial in Adults (CERTA)

Impact on Clinical Outcome of Continuous Video-electroencephalography (cEEG) Monitoring in Patients With Disorders of Consciousness: A Randomized Controlled Trial

Continuous video-EEG monitoring (cEEG) significantly improves seizure or status epilepticus detection in patients in intensive care units (ICUs), and is recommended for patients with consciousness impairment. cEEG is time- and resource consuming as compared to routine EEG (rEEG, lasting 20-30 minutes). While centers in North America have been using it increasingly, most European hospitals still do not have resources to comply with these guidelines. In addition, only one population-based study based on discharge diagnoses suggested that cEEG may improve patients' outcome. Current guidelines are thus based upon weak evidence and expert opinions. Aim of the study is to assess if cEEG in adults with consciousness impairment is related to an improvement of functional outcome, and to address the prognostic role of quantitative network EEG analyses. In this multicenter randomized controlled trial, adults with GCS inferior or equal to 11 or FOUR score inferior or equal to 12 will be randomized 1:1 to cEEG for 30-48 hours or two rEEG within 48 hours. The primary outcome will be mortality at 6 months. Secondary outcomes will blindly assess functional outcome, seizure/status epilepticus detection rate, duration of ICU stay, change in patient management (antiepileptic drug introduced, increased, or stopped, brain imaging), and reimbursement. Additionally, quantitative EEG will be assessed towards the primary outcome. 350 patients are planned to be included.

Background: Continuous video-EEG monitoring (cEEG) is a non-invasive tool to monitor the electrical brain function; it significantly improves seizure or status epilepticus detection in comatose patients in intensive care units (ICUs), which often do not show any specific clinical correlates. Recently, the European Society of Intensive Care Medicine published guidelines regarding the use of cEEG in the ICUs, recommending it for most patients with consciousness disorders. cEEG is time- and resource consuming as compared to routine spot EEG (rEEG, typically lasting 20-30 minutes). While centers in North America have been using it increasingly, most European - and all Swiss - hospitals still do not have enough resources to comply with these guidelines. In addition, while the superiority of cEEG to detect non-convulsive seizures or status epilepticus is proven, only one population-based study based on discharge diagnoses suggested that cEEG may improve patients' outcome. Current guidelines are thus based upon weak evidence and expert opinions. If cEEG leads to improved patients' care remains elusive. Moreover, little attention has been drawn towards quantitative EEG information beyond visual analysis, and the impact of such information on diagnosis, treatment, and outcome remains unclear. Aim: To assess whether the use of cEEG in patients with consciousness impairment is related to an improvement of functional outcome, and to address the prognostic role of quantitative network EEG analyses in this cohort. Also, a cost analysis will be performed. Methods: In this multicenter randomized controlled trial, adults with a Glasgow Coma Score (GCS) inferior or equal to 11 or a FOUR score inferior or equal to 12, regardless of etiologies, will be randomized 1:1 to cEEG for 30-48 hours or two rEEG within 48 hours, interpreted in a standardized way. Patients with detected seizures in the last 36h or status epilepticus in the last 96h will be excluded, as cEEG may represent the standard of care. Demographics, etiology, Charlson Comorbidity Index, GCS, diagnosis leading to EEG, mechanical ventilation, and subsequent use of rEEG/cEEG will be collected. The primary outcome will be mortality at 6 months. Secondary outcomes will blindly assess functional outcome at 4 weeks and 6 months, as well as seizure/status epilepticus detection rate and time to detection, infections rate, duration of ICU stay, change in patient management (antiepileptic drug introduced, increased, or stopped, brain imaging), and reimbursement. Analyses will compare the two interventional groups (intention to diagnose) regarding outcome, as a whole and stratified according to etiological subgroups, and other variables of interest. Additionally, lope cross correlation and horizontal visibility graphs will be applied to compute a weighted adjacency matrix consisting of all the pairwise interdependences between EEG signals, in order to characterize the integrative and segregative characteristics of the underlying functional brain networks and compare their relationship with the primary outcome. According to a previous estimate, patients with consciousness disorders undergoing cEEG have a 75% survival rate; while patients w/o cEEG 61%. Using a power of 0.8, an α error of 0.05, and a 2-side approach, 2x174 patients would be needed to detect this significant difference in survival. Expected impact: This study will clarify if cEEG monitoring has a significant impact on functional outcome and define its cost effectiveness, and if network EEG analysis has a role in outcome prognostication. The results of this study will have a considerable potential to influence clinical practice regarding EEG and treatment of patients with altered levels of consciousness. If results will indicate that cEEG contributes to improve outcome, this will lead to the urgent need for implementation of cEEG with consecutive substantial impact on health care and resource allocation in larger Swiss and European hospitals.

Patients randomized to continuous EEG will be recorded with at least 21 electrodes placed according to the international 10-20 system; occasionally, a reduced montage will be allowed in patients with extensive neurosurgical scars, according to good common practice. Recordings will last a minimum of 30 and a maximum of 48 hours. During this time, one interruption to a maximum of two hours for diagnostic purposes will be allowed. Reactivity testing using auditory and nociceptive stimuli will be performed at least twice during the recording time. Recordings will be visually interpreted by certified electroencephalographers (i.e., interpretation of the automated algorithm only won't be allowed) using the 2013 American Clinical neurophysiology nomenclature; interpretations will be communicated within two hours of their completion to the treating team.

Patients randomized to routine EEG will be recorded with at least 21 electrodes placed according to the international 10-20 system; occasionally, a reduced montage will be allowed in patients with extensive neurosurgical scars, according to good common practice. Recordings will last between 20 and 30 minutes; two recordings will take place over a period of 24 to 48 hours. Reactivity testing using auditory and nociceptive stimuli will be performed once per recording. Recordings will be visually interpreted by certified electroencephalographers using the 2013 American Clinical neurophysiology nomenclature, as for the experimental intervention, and the interpretation will be communicated within two hours of its completion to the treating team.

Rate of in-hospital infections requiring antibiotic treatment at 4 weeks after first EEG (proportion)

Change in clinical patient management (i.e., antiepileptic drugs (AED) introduced or stopped, AED increased or decreased, brain imaging procedure order) occurring during the 60 hours following the start of the first EEG (categorical).

Global hospitalization costs intended as amount billed for each patient's acute hospital stay, assessed through the billing department of each hospital (continuous variable - stratified by site)

Inclusion Criteria: In-patients aged ≥18 years, treated in an ICU or intermediate care unit Alteration of mental state of any etiology (i.e., primarily cerebral or not), with Glasgow-coma scale inferior or equal to 11 or FOUR score inferior or equal to 12. Need of an EEG to exclude seizures or SE, or to evaluate prognosis as per the treating physician or the consulting neurologist. Informed consent obtained for research in emergency situation according to Human Research Act (HRA) art 30-31 at the time of inclusion Exclusion Criteria: Clinical and/or electrographic status epilepticus < 96h before randomization Clinical and/or electrographic seizure < 36h before randomization Palliative care situation, in which detection of SE or seizures would not have any impact on the patient's care. High likelihood of needing a surgical intervention or an invasive diagnostic procedure within the next 48 hours according to the treating physician (as this would require cEEG removal).

Urbano V, Alvarez V, Schindler K, Ruegg S, Ben-Hamouda N, Novy J, Rossetti AO. Continuous versus routine EEG in patients after cardiac arrest: Analysis of a randomized controlled trial (CERTA). Resuscitation. 2022 Jul;176:68-73. doi: 10.1016/j.resuscitation.2022.05.017. Epub 2022 May 30.

Urbano V, Novy J, Alvarez V, Schindler K, Ruegg S, Rossetti AO. EEG recording latency in critically ill patients: Impact on outcome. An analysis of a randomized controlled trial (CERTA). Clin Neurophysiol. 2022 Jul;139:23-27. doi: 10.1016/j.clinph.2022.04.003. Epub 2022 Apr 18.

Urbano V, Novy J, Schindler K, Ruegg S, Alvarez V, Zubler F, Oddo M, Lee JW, Rossetti AO. Continuous versus routine EEG in critically ill adults: reimbursement analysis of a randomised trial. Swiss Med Wkly. 2021 Mar 16;151:w20477. doi: 10.4414/smw.2021.20477. eCollection 2021 Mar 15.

Guinchard M, Warpelin-Decrausaz L, Schindler K, Ruegg S, Oddo M, Novy J, Alvarez V, Rossetti AO. Informed consent in critically ill adults participating to a randomized trial. Brain Behav. 2021 Feb;11(2):e01965. doi: 10.1002/brb3.1965. Epub 2020 Dec 3.

Rossetti AO, Schindler K, Sutter R, Ruegg S, Zubler F, Novy J, Oddo M, Warpelin-Decrausaz L, Alvarez V. Continuous vs Routine Electroencephalogram in Critically Ill Adults With Altered Consciousness and No Recent Seizure: A Multicenter Randomized Clinical Trial. JAMA Neurol. 2020 Oct 1;77(10):1225-1232. doi: 10.1001/jamaneurol.2020.2264.