Comatose patients have varying degrees of neurological failure, which is the main cause of mortality and morbidity in intensive care units(28). Due to its easy operation, the GCS score is frequently used in clinics to evaluate the neurological failure of coma patients(29). But the results of this study showed a poor correlation between GCS score and prognosis. EEG-R is a new method for clinicians to predict the prognosis of patients with coma patients. The integrity of neurological pathways could be reflected by EEG-R. The EEG is nonreactive when severe cerebral dysfunction prevents impulses from peripheral stimuli from reaching the cerebral cortex. In critically ill patients with diminished consciousness, a lack of EEG-R was associated with mortality(3, 30, 31). Numerous studies investigated the prognostic value of EEG-R in the past decade. However, these studies varied in the results and no studies were evaluated comprehensively.
This systematic review included 12 studies and is the first to comprehensively analyze the prognostic value of EEG-R. We found that EEG-R has high accuracy to predict the prognosis of coma patients, in terms of both sensitivity and specificity. Concerning the impact of the etiology and prognostic assessment criteria, we furtherly conducted subgroup analyses. The subgroup analyses indicated that results are stable whatever the methods used in prognosis assessment and Whatever the etiology of patients is.
Our study is the largest and most up-to-date systematic review in terms of the prognostic aspects of EEG-R. Consistent with the findings of Taggar et al. [38], both sensitivity and specificity were particularly profound. Therefore, EEG-R is a useful prognostic indicator and may shift the treatment of critically ill patients. Further studies are encouraged to assess the predicted value based on EEG.
Notably, the study has some limitations. The substantial heterogeneity in our study was inevitable. Firstly, Most studies do not mention the exact time EEG was evaluated. EEG features may change during the acute stage, especially in the first 24–48 hours after a coma(32). However, there are five studies included in our analysis that performed EEG monitor before 48 hours, which may reflect the false condition of patients. Repeated continuous EEG monitoring after 48 hours is then recommended to assess reactivity in future research.
Secondly, the included studies used different external stimuli. The external stimuli applicated in patients with impaired consciousness conventionally consist of: auditory and/or nociceptive stimulation (33), as well as, more rarely, passive eye-opening and photic stimulation(34, 35). There is also no standardized intensity regarding stimulation. Consistent stimulation methods are needed.
Finally, current practice of evaluating EEG-R though visual analysis generated inter-observer variation. It is diffcult for researchers to judge reative or nonreactive from subtle changes without special training. These subtle changes ingored by researchers may impair overall prognostic value. TH and sedative drugs also lead to difficult visual assessment (36).