This study investigates the prognostic value of EEG reactivity, the consistency of sensory stimulation, and inter-rater variability in comatose intensive care unit (ICU) patients. EEG reactivity is a key component of neuro-prognostication following hypoxic–ischaemic brain injury; however, assessments remain highly subjective and lack standardisation across centres. Variability in stimulation techniques and interpretation can limit the reliability and clinical utility of EEG reactivity findings.

A retrospective audit was conducted at the Queen Elizabeth University Hospital, Glasgow, analysing EEG recordings from 83 ICU patients between April and July 2024. Data were reviewed to evaluate the types of stimulation used during recordings, their documentation in EEG reports, and inter-rater agreement among Clinical Physiologists and Neurophysiologists when classifying EEG reactivity. Statistical analysis included a Pearson’s Chi-Square test to assess the relationship between EEG reactivity and survival outcomes, and Fleiss’ Kappa to evaluate inter-rater reliability.

Results showed that 53% of patients exhibited no EEG reactivity, while 16.9% demonstrated reactivity. The remaining patients displayed uncertain or ambiguous responses. No significant association was found between EEG reactivity and survival (p = 0.533), suggesting that EEG reactivity alone is not a reliable prognostic marker in this cohort. Inter-rater variability was moderate (Fleiss’ Kappa = 0.512), reflecting variability in interpretation even among experienced raters. Stimulation methods were inconsistently applied, with 17.5% of EEG reports lacking documentation regarding the type or intensity of stimuli. In particular, noxious stimulation was inconsistently performed in unreactive patients, potentially influencing overall assessment accuracy.

These findings emphasise the need for a structured and standardised EEG reactivity protocol to improve consistency, reproducibility, and clinical decision-making in critical care. The development of clear stimulation guidelines, improved documentation standards, and incorporation of quantitative EEG measures could reduce subjectivity and enhance prognostic accuracy. Future research should explore multi-centre collaboration to establish consensus-based reactivity testing procedures and to validate the prognostic role of EEG reactivity using objective metrics. This audit was registered within the hospital’s clinical governance framework and supports ongoing efforts to standardise neurophysiological practice in the ICU setting.