As resuscitation and life support techniques improve, more and more patients are surviving, but with varying degrees of brain damage, especially after cardiopulmonary resuscitation (CPR). Less than 20% of patients survive in-hospital CPR until discharge, and of those patients, 80% experience coma of varying length and 40% enter a persistent vegetative state with low chances of full recovery of brain function. Clinicians must consider the timing of limiting or withdrawing life support. Therefore, early determination of neurological prognosis is important, and accurate and reliable indicators must be used to assess the prognosis. Clinical assessment Clinical situation A thorough history helps to determine the etiology and prognosis, which includes: (i) time and speed of onset and prodromal symptoms, (ii) history of disease and medication, and (iii) time to resuscitation and time to cerebral ischemia, but none of these factors is sufficiently reliable to determine the prognosis of brain function. The relevant conditions have some reference value, including: time of hypoxia, CPR duration, time of cardiac arrest, and type of arrhythmia. The prognosis of cardiac arrest due to ventricular arrhythmias or tachycardia is often better when there is a witness at the time of arrest; some studies have shown that out-of-hospital CPR does not exceed 15 min in those who survive more than 6 weeks, but it cannot be used as a prognostic criterion: the false positive rate (FPR) for these parameters is as high as 20%-27%. Other factors associated with poor prognosis include: temperature, age >70 years, pre-admission stroke, renal failure and congestive heart failure. Neurological assessment Neurological examination is the most widely studied parameter for predicting poor outcome in coma survivors of cardiac arrest. It includes the presence of voluntary movements, response to sound, light, pain stimuli, pupil size and reflex to light, other cranial nerve functions including corneal and cranio-ocular reflexes, type of respiration (voluntary, rhythm disturbance, etc.). The prognostic criteria of the Glasgow Coma Scale (GCS) were: GCS ≥ 13 points, indicating mild brain injury; GCS between 9 and 12 points, indicating moderate brain injury; GCS ≤ 8 points, indicating severe brain injury; GCS ≤ 4 points in the first 48 h, indicating poor prognosis (death, persistent coma). Myoclonic persistent epilepsy (MSE) is usually associated with in-hospital death or poor prognosis (even if the patient has intact brainstem reflexes and partial motor reflexes). However, it has been suggested that MSE alone is not a reliable predictor of poor prognosis and that only a few patients recover well, so a combination of other indicators is needed. Absence of pupil-to-light reflex and corneal reflex 24 h after resuscitation and no motor response 24-72 h after resuscitation suggest poor prognosis (severe neurological disability or death). The specificity of poor prognosis was 100% in the following two cases: (i) 72 h post-resuscitation motor response was denervation or no response (FPR=0), and (ii) 72 h post-resuscitation pupil-to-light reflex and corneal reflex were absent (FPR=0). Confounding factors for prognostic judgment include: medications (e.g., anticholinergics, sedatives, neuromuscular blockers), acute metabolic disturbances, especially hepatic and renal failure, and shock. Electrophysiological monitoring EEG The value of EEG for assessing hypoxic brain injury has not been established due to the different classification systems and recording intervals used by different investigators, the subjective nature of EEG interpretation, and the susceptibility to metabolic disturbances, sedative drugs, and sepsis. The EEG of ischemic-hypoxic brain injury is divided into two main categories: (1) malignant: including complete or near-complete depression, fulminant depression, extensive periodic complex waves, low voltage waves (≤10 μV), intermittent or continuous epileptic waves, unresponsiveness to stimulation and alpha-theta waves; (2) benign: the presence of variability and reactivity. It has been shown that the presence of malignant waves is associated with higher mortality and poorer prognosis. Abnormal evoked potentials are associated with poor outcome. Lack of bilateral N20 waves in median somatosensory evoked potentials (SSEP) stimulated 24-72 h after resuscitation is associated with poor prognosis, but the sensitivity is only 46% , and the sensitivity of other evoked potentials is even lower. Blood and cerebrospinal fluid biomarkers Biomarkers are released by dying nerves or glial cells and are detectable in blood or cerebrospinal fluid, and their prognostic criteria are: neuron-specific enolase (NSE) (currently the most meaningful marker of brain injury) >33 μg/L, S100B >1.2 μg/L, glial acidic protein (GFAP) >0.1 ng/ dL, and serum brain-type creatine kinase (CKBB) >10 U/L. The most important advantage of these criteria is that their levels are not interfered with by sedation or neuromuscular blocking agents. Neuroimaging Many post-resuscitation functional magnetic resonance imaging (MRI) studies have shown that extensive cortical and subcortical damage is associated with poor neurological prognosis, and MRI helps to differentiate between a vegetative state and a minimally conscious state. Many computed tomography (CT) parameters (including quantitative gray matter-white matter or Sphere’s unit ratios and qualitative description of brain structures) are also associated with poor prognosis, and plain CT scans can also provide information on brain injury, cerebral infarction, intracranial hemorrhage, and other conditions that may lead to cardiac arrest. Other parameters such as positron emission tomography (PET) or CT, cerebral angiography, and transcranial Doppler ultrasound (TCD) are less well studied. Despite its potential value, neurological imaging has not been shown to be an independent predictor of poor prognosis, and it requires removal of the patient from the custodial setting and is not indicated in certain critically ill patients. Hypothermia Over the past 10 years, 2 randomized controlled trials have demonstrated that subhypothermia (32-34°C) for 24-48 h significantly improves neurological prognosis in comatose patients who have resumed spontaneous respiratory circulation after cardiac arrest due to ventricular fibrillation. Subcryotherapy is recommended by the International Society of Resuscitation and the American Heart Association for the treatment of coma after CPR. Hypothermia studies have shown that prognostic indicators are affected by sedation and neuromuscular blocking agents, with increased false positive rates, and are no longer effective in predicting the prognosis of patients treated with hypothermia. Patients receiving hypothermia after restoration of autonomic circulation need to be continuously observed for more than 72 h before their poor prognosis can be evaluated. Hypothermia improves the prognosis but also changes the value of neurological examination to determine prognosis. Early EEG examination and active control of electroencephalographic epilepsy are recommended. After restoration of normal body temperature for 48-72 h, NSE and SSEP examination, CT examination if necessary, clinical examination after 72 h, and MRI examination after 3-5 d. It is recommended to integrate all clinical neurological examinations, EEG, SSEP, and biomarkers for judgment. In conclusion, resuscitation such as cardiac arrest may be complicated by ischemic-hypoxic brain injury and coma at an early stage, and its prognosis can be awakening, persistent vegetative state or brain death. The absence of pupil-to-light reflex and corneal reflex in the first 3 d after cardiac resuscitation suggests a poor prognosis. In patients treated with subhypothermia, the prognosis is not reliably predicted based on the motor response index at day 3. The lack of SSEP bilateral N20 wave was associated with poor prognosis 24-72 h after resuscitation, but the sensitivity was poor; NSE was a meaningful biomarker for predicting poor prognosis, but the shortcoming was that the prediction was unstable in patients with subhypothermia, and further studies were needed; 24 h after resuscitation, both SSEP and NSE could predict poor prognosis, which was significant; potential confounding factors for prognosis were: acute metabolic potential confounding factors for prognostic judgment are: acute metabolic disorders (e.g., liver and kidney failure, shock), sedatives and neuromuscular blocking agents, and subhypothermia treatment should be continued if there are no adverse prognostic indicators.