EEG Signal Correlates During Anesthesia and Sedation in Adults and Children

Overview

This prospective, multicenter, non-interventional observational study investigates the correlates between electroencephalographic (EEG) signals and pharmacological/clinical parameters during general anesthesia and procedural sedation. The study will enroll 330 participants across three distinct populations at two AP-HP sites (Bichat-Claude Bernard Hospital and Louis Mourier Hospital): 1. Pediatric cohort (n=110): Children aged 1-17 years undergoing general anesthesia, including 55 with autism spectrum disorder (ASD) and 55 without ASD. 2. Elderly adult cohort (n=110): Patients over 70 years undergoing scheduled surgery under general anesthesia. 3. Procedural sedation cohort (n=110): Adults ≥18 years undergoing procedural sedation for digestive endoscopy. EEG data will be collected using BIS and SedLine monitors during routine anesthetic care with no modifications to standard practice. The primary objective is to identify EEG signal features correlating with clinical and pharmacological parameters during induction, maintenance, and emergence phases. Secondary objectives include developing predictive models for anesthetic depth, analyzing age-related differences, comparing effects of different anesthetic agents, and investigating specific EEG patterns in children with ASD. This study will enhance understanding of brain responses to anesthesia across different age groups and clinical contexts, potentially improving anesthetic monitoring algorithms and management strategies.

BACKGROUND AND RATIONALE: Monitoring depth of anesthesia using electroencephalography (EEG) has become standard clinical practice, particularly through processed indices like the Bispectral Index (BIS). However, the underlying neurophysiological mechanisms and relationships between raw EEG signals and clinical/pharmacological parameters remain incompletely understood. Current monitoring systems provide processed indices but limited insight into raw multi-channel EEG signal characteristics. Systematic analysis of EEG data across different patient populations, age groups, and anesthetic contexts is needed to advance our understanding of brain responses to anesthesia and improve monitoring strategies. STUDY DESIGN: This is a prospective, multicenter, non-interventional, observational cohort study conducted at two sites of the Assistance Publique - Hôpitaux de Paris (AP-HP): Bichat-Claude Bernard Hospital and Louis Mourier Hospital. The study involves no modifications to routine anesthetic care - only collection and analysis of EEG data during standard clinical practice. STUDY POPULATIONS: The study will enroll 330 participants across three distinct cohorts: Cohort 1 - Pediatric Population (n=110): Children aged 1-17 years undergoing general anesthesia for scheduled surgical procedures. This cohort specifically includes: * 55 children with autism spectrum disorder (ASD) diagnosed according to DSM-5 criteria * 55 children without ASD (matched controls) This unique population allows investigation of potential neurodevelopmental differences in anesthetic response and EEG patterns in children with ASD. Cohort 2 - Elderly Adult Population (n=110): Patients aged \>70 years undergoing scheduled surgery under general anesthesia at Bichat-Claude Bernard Hospital. This elderly population is important for studying age-related physiological changes affecting EEG patterns during anesthesia and potential increased sensitivity to anesthetic agents. Cohort 3 - Procedural Sedation Population (n=110): Adults (≥18 years) undergoing procedural sedation for diagnostic or therapeutic digestive endoscopy procedures at Louis Mourier Hospital. This cohort provides comparative data on lighter levels of sedation versus general anesthesia. INCLUSION CRITERIA: For adult patients: * Age ≥18 years (Cohort 2: \>70 years; Cohort 3: ≥18 years) * Scheduled for general anesthesia (Cohort 2) or procedural sedation (Cohort 3) * Pre-anesthetic consultation completed * Informed of the study and non-opposition obtained For pediatric patients (Cohort 1): * Age 1-17 years * Scheduled for general anesthesia for surgical procedure * For ASD subgroup: Confirmed DSM-5 diagnosis of autism spectrum disorder * Parental authority holder informed and non-opposition obtained * Child informed (if age-appropriate) and non-opposition obtained EXCLUSION CRITERIA: For adult patients: * Documented cognitive impairment in medical record * Patient refusal to participate * Patient deprived of liberty or under legal protection * Pregnant or breastfeeding women For pediatric patients: * Parental refusal of child's participation * Child refusal (if age-appropriate to understand) DATA COLLECTION: During routine anesthetic care without any protocol-specific interventions, the following will be recorded: EEG Monitoring: * Continuous multi-channel EEG using BIS (Bispectral Index) monitor * Continuous multi-channel EEG using SedLine monitor * Processed parameters (BIS index, spectral edge frequency, suppression ratio, etc.) * Raw EEG signals recorded in standard EDF format Synchronized Clinical Annotations: * Anesthetic drug administration (agent, dose, route, timing) * Clinical events (intubation, surgical incision, emergence, etc.) * Hemodynamic parameters (heart rate, blood pressure) * Level of consciousness assessments (RASS, clinical signs) * Surgical stimulation intensity * Adverse events Patient Characteristics: * Demographics (age, sex, weight, height, BMI) * Medical history and comorbidities * Concomitant medications * ASA physical status classification * For pediatric cohort: ASD diagnosis and characteristics All EEG recordings and multi-signal files will be pseudonymized using a unique study identifier before analysis. Data will be stored on secure servers compliant with GDPR regulations. PRIMARY OBJECTIVE: To identify features of the electroencephalographic (EEG) signal that significantly correlate with: 1. Pharmacological parameters: anesthetic drug type, concentration (estimated effect-site concentration), drug combinations, pharmacokinetic/pharmacodynamic properties 2. Clinical parameters: patient age, comorbidities, surgical stimulation intensity, hemodynamic changes, emergence characteristics These correlations will be analyzed separately for the induction, maintenance, and emergence phases of general anesthesia and for procedural sedation. SECONDARY OBJECTIVES: 1. Develop and validate predictive models using machine learning algorithms to link specific EEG features to anesthetic depth and level of consciousness 2. Compare EEG spectral and temporal characteristics between pediatric and adult populations during equivalent anesthetic states 3. Analyze differential effects of commonly used anesthetic agents (propofol, sevoflurane, desflurane, remifentanil) on multi-channel EEG patterns 4. Investigate specific EEG signatures and anesthetic responses in children with autism spectrum disorder compared to typically developing children 5. Characterize distinctive EEG patterns during procedural sedation versus general anesthesia 6. Evaluate inter-individual and intra-individual variability in EEG responses to anesthetic agents 7. Assess correlation between frontal EEG patterns and clinical signs of anesthetic depth ENDPOINTS: Primary Endpoint: Correlation coefficients (Pearson or Spearman as appropriate) between extracted EEG features (spectral power in different frequency bands, coherence, entropy measures, burst suppression patterns) and pharmacological/clinical variables during different anesthetic phases. Secondary Endpoints: * Performance metrics (sensitivity, specificity, AUC) of predictive models for anesthetic depth * Differences in EEG power spectral density between age groups * Changes in EEG patterns associated with specific anesthetic agents * Differences in EEG characteristics between children with/without ASD * Comparison of EEG patterns between sedation and general anesthesia * Variance components of inter- and intra-individual EEG variability STATISTICAL ANALYSIS: Sample size of 330 participants (110 per cohort) was calculated to provide: * Adequate power (\>80%) to detect medium effect sizes (r≥0.3) for correlation analyses * Sufficient training and validation datasets for machine learning models * Subgroup analyses within pediatric cohort (ASD vs non-ASD) Advanced signal processing techniques will be applied including: * Fast Fourier Transform for spectral analysis * Wavelet decomposition for time-frequency analysis * Entropy measures (spectral entropy, permutation entropy) * Coherence and connectivity analyses * Machine learning algorithms (random forests, support vector machines, neural networks) Statistical analyses will use appropriate parametric or non-parametric tests depending on data distribution. Multiple comparison corrections will be applied when appropriate (Bonferroni, FDR). Mixed-effects models will account for repeated measures within individuals. STUDY DURATION: * Recruitment period: 12 months * Individual patient participation: Approximately 1 month (interval between pre-anesthetic consultation and surgical procedure/endoscopy) * Total study duration: 13 months including data analysis ETHICAL AND REGULATORY CONSIDERATIONS: This non-interventional study involves no modifications to standard clinical care, no additional risk to participants, and no experimental interventions. The study has been submitted to the Comité de Protection des Personnes (CPP) for ethical approval and will be registered with the ANSM. Participants (or parents for minors) will receive written information and provide non-opposition for data use. All data management complies with GDPR regulations (MR-003 reference methodology). EXPECTED OUTCOMES AND IMPACT: This comprehensive study will provide rich datasets on EEG signal characteristics across diverse anesthetic contexts and patient populations. Results may lead to: * Improved algorithms for real-time anesthetic depth monitoring * Better understanding of age-related and neurodevelopmental differences in anesthetic response * Optimization of anesthetic agent selection and dosing strategies * Enhanced safety through more accurate consciousness monitoring * New insights into neurophysiological mechanisms of anesthesia * Foundation for future development of personalized anesthetic management approaches The purely observational design ensures maximal safety while generating clinically relevant data to advance the field of neuromonitoring in anesthesia.