Version July 2025
INTRODUCTION —
Neonatal encephalopathy is a heterogeneous, clinically defined syndrome characterized by disturbed neurologic function in the earliest days of life in an infant born at or beyond 35 weeks of gestation, manifested by a reduced level of consciousness or seizures. It is often accompanied by difficulty with initiating and maintaining respiration and by depression of tone and reflexes [1].
This topic will review the treatment and prognosis of neonatal encephalopathy. Other aspects are discussed separately. (See "Neonatal encephalopathy: Etiology and pathogenesis" and "Neonatal encephalopathy: Clinical features and diagnosis".)
Related topics include:
●(See "Perinatal asphyxia in term and late preterm infants".)
●(See "Clinical features, evaluation, and diagnosis of neonatal seizures".)
●(See "Treatment of neonatal seizures".)
●(See "Etiology and prognosis of neonatal seizures".)
TREATMENT —
Management of neonatal encephalopathy consists of supportive care for all patients and therapeutic hypothermia for those who meet criteria (table 1). (See 'Therapeutic hypothermia' below and 'Supportive management' below.)
Supportive management — Neonates with moderate and severe neonatal encephalopathy should be cared for in the neonatal intensive care unit (NICU) setting. Major goals of supportive care include the maintenance of physiologic homeostasis and treatment of overt manifestations of brain injury [2,3].
Key aspects of supportive care include the following, each of which is discussed in separate topic reviews:
●Maintain adequate ventilation – Hypoxemia, hyperoxia, and/or respiratory acidosis should be avoided. Most neonates with moderate or severe hypoxic-ischemic encephalopathy (HIE) who are managed with therapeutic hypothermia require endotracheal intubation and mechanical ventilation. Neonates with severe asphyxia may develop persistent pulmonary hypertension (PPHN) and may require advanced respiratory care (eg, high-frequency ventilation, inhaled nitric oxide, and/or extracorporeal membrane oxygenation [ECMO]). While therapeutic hypothermia during ECMO may increase risk for intracranial hemorrhage, therapeutic hypothermia is not considered to be a contraindication for ECMO [4,5]. (See "Overview of mechanical ventilation in neonates" and "Persistent pulmonary hypertension of the newborn (PPHN): Management and outcome".)
●Maintain adequate systemic and cerebral perfusion – Inotropic agents may be required to maintain blood pressure. However, systemic hypertension and volume overload should be avoided since they can worsen cerebral edema. Cerebral near infrared spectroscopy (NIRS) can provide continuous real-time information about changes in brain hemodynamics, oxygenation, and metabolism [6]. However, the utility of NIRS in guiding therapeutic interventions in neonatal encephalopathy is not established. Steadily increasing cerebral NIRS reading above an infant's baseline and a value of >90 percent may indicate significant neurologic injury [7]. (See "Neonatal shock: Management", section on 'Vasoactive agents'.)
●Maintain normal fluid, electrolyte, and acid-base status – Avoid fluid overload, provide adequate nutrition, and avoid acidosis. (See "Fluid and electrolyte therapy in newborns" and "Parenteral nutrition in infants and children".)
●Maintain euglycemia – Hypoglycemia and hyperglycemia in infants with moderate to severe neonatal encephalopathy are associated with adverse outcomes [8,9]. (See "Management and outcome of neonatal hypoglycemia" and "Neonatal hyperglycemia".)
●Control seizures – All infants with neonatal encephalopathy are considered to be at high risk for seizures. Continuous electroencephalography (cEEG) is the optimal method for detection and diagnosis; amplitude-integrated EEG (aEEG) is an alternative if cEEG is not available [10]. Phenobarbital is considered a first-line antiseizure medication for most etiologies of neonatal seizures, including HIE. (See "Neonatal encephalopathy: Clinical features and diagnosis", section on 'Electroencephalography' and "Treatment of neonatal seizures".)
●Consult a metabolic/genetic specialist if metabolic disorder suspected – If a metabolic disorder is suspected as the underlying etiology of encephalopathy, urgent consultation with a specialist in pediatric metabolic disorders or genetics should be sought, and feeds should be held pending the evaluation. (See "Metabolic emergencies in suspected inborn errors of metabolism: Presentation, evaluation, and management" and "Inborn errors of metabolism: Identifying the specific disorder".)
General supportive measures for infants with a perinatal hypoxic-ischemic event are reviewed in detail separately. (See "Perinatal asphyxia in term and late preterm infants", section on 'Supportive care based on organ system'.)
Therapeutic hypothermia — Therapeutic hypothermia is the only proven neuroprotective therapy for treatment of neonatal encephalopathy and is available in most experienced neonatal centers. Therapeutic hypothermia is started within the first six hours after delivery and maintained for 72 hours at 33 to 35°C (91.4 to 95°F).
The efficacy of therapeutic hypothermia is supported by randomized trials and meta-analyses, which are discussed below. (See 'Effectiveness' below.)
Indications — For term or late preterm infants with moderate to severe neonatal HIE (table 2) who are managed at experienced centers capable of providing comprehensive care, we recommend therapeutic hypothermia with head or whole-body cooling starting in the first six hours after birth [11]. Implementation of therapeutic hypothermia should follow protocols used in one of the major published trials. (See 'Implementation' below.)
Eligibility criteria for therapeutic hypothermia are as follows (table 1) [11]:
●Gestational age ≥36 weeks and ≤6 hours of age. Some centers include gestational age ≥34 or 35 weeks, based on limited data [12,13].
●One or more of the following:
•Metabolic or mixed acidosis with a pH of ≤7.0 or a base deficit of ≥16 mmol/L in a sample of umbilical cord blood or any blood obtained within the first hour after birth
•A 10-minute Apgar score of ≤5
•Ongoing resuscitation (eg, assisted ventilation, chest compressions, or cardiac medications) initiated at birth and continued for at least 10 minutes
●Moderate to severe encephalopathy on clinical examination. There is practice variation regarding how this criterion is defined [14]. Most centers use a modified Sarnat examination (table 2), with or without additional information on the presence of seizures. The Sarnat examination rates the severity of abnormalities in level of consciousness, spontaneous activity, tone, posture, primitive reflexes, and autonomic function [15].
There is a consensus among experts that therapeutic hypothermia should be more widely available based upon the benefit and safety of hypothermia and the lack of other effective treatments [16-18]. Thus, hypothermia has become the standard of care in most NICUs in the United States, Canada, Europe, Australia, and Japan, and guidelines support the use of therapeutic hypothermia for infants who meet the criteria used in the published trials [19-22]. (See 'Society guideline links' below.)
Implementation — Hypothermia should be started within the first six hours after delivery and continued for 72 hours at target temperature (table 1). The rectal temperature should be maintained at 33 to 35°C (91.4 to 95.0°F), with target temperature typically set at 33.5°C. Although direct comparisons are lacking, selective head cooling and whole-body cooling appear to have similar safety and effectiveness. However, whole-body cooling is preferred to head cooling in most centers in the United States due to ease of administration. Whole-body cooling also provides easier access to the scalp for electroencephalography (EEG) monitoring.
Cooling can be started and maintained during neonatal transport if there is need to transfer the infant to a specialized center [23-25]. Neonates managed with therapeutic hypothermia may require analgesia and/or sedation if there are signs of distress or shivering. (See "Management and prevention of pain in neonates", section on 'Hypoxic-ischemic encephalopathy'.)
Hypothermia for 72 hours with the rectal temperature maintained at 33 to 35°C (91.4 to 95.0°F) is the strategy used by the major randomized trials that established the efficacy of this intervention. Four trials (NICHD [26], TOBY [27], neo.nEURO.network [28], and ICE [29]) used whole-body cooling to a goal rectal temperature of 33.5°C (92.5°F). Three trials (CoolCap [30] and two others [31,32]) employed selective head cooling, with a target rectal temperature 34 to 35°C (93.2 to 95.0°F).
Longer duration of cooling does not seem to confer additional benefit. A subsequent clinical trial compared the standard cooling protocol (72 hours at 33.5°C [92.3°F]) versus longer (120 hours) or deeper (32°C [89.6°F]) hypothermia, or both [33,34]. The trial was stopped early for futility and a signal of possible harm in the experimental arms [33]. At mean follow-up of 21 months, rates of death or disability were similar in patients who received 72 hours of cooling compared with those who received 120 hours of cooling (32 percent in each group). Rates of death or disability were also similar in patients who received hypothermia to 33.5°C versus 32°C (32 percent in each group).
Adverse effects — Therapeutic hypothermia is generally well tolerated. Short-term adverse effects in the randomized trials with therapeutic hypothermia compared with standard care included increased rates of sinus bradycardia (9.6 versus 0.5 percent; risk ratio [RR] 11.6, 95% CI 4.9-27.2) and thrombocytopenia (35 versus 28 percent; RR 1.21, 95% CI 1.2-1.4) [35]. However, rates of cardiac arrhythmia requiring medical intervention were quite low in both therapeutic hypothermia and standard care (0.3 and 0.6 percent, respectively).
Subcutaneous fat necrosis has been observed as a potential complication of therapeutic hypothermia, affecting <1 to 3 percent of treated neonates [36-40]. Onset can occur days to weeks after cooling and may be accompanied by hypercalcemia, hypoglycemia, and/or thrombocytopenia. Subcutaneous fat necrosis is generally benign and self-limited, but hypercalcemia can cause potentially life-threatening symptoms requiring treatment. The clinical manifestations, diagnosis, and management of this disorder are reviewed in detail elsewhere. (See "Subcutaneous fat necrosis of the newborn".)
Another adverse effect of therapeutic hypothermia is delayed bonding between parents and their newborn during the period of hypothermia [41].
Effectiveness
For moderate to severe hypoxic-ischemic encephalopathy
Benefit in high-income countries — Clinical trials performed in high-income countries have demonstrated that therapeutic hypothermia improves survival and neurodevelopmental outcomes for infants with moderate to severe HIE [35,42-44]. In a meta-analysis with data from eight studies and 1344 neonates with moderate to severe HIE who were randomly assigned to therapeutic hypothermia (starting within six hours after birth) or standard care, therapeutic hypothermia reduced the combined outcome of mortality or major neurodevelopmental disability at 18 months of age (RR 0.75, 95% CI 0.68-0.83; risk difference [RD] ‐15 percent, 95% CI ‐20 to ‐10 percent; number needed to treat (NNT) for an additional beneficial outcome 7, 95% CI 5-10) [35].
●Improved survival – In a meta-analysis of 10 trials including 1468 neonates with moderate to severe HIE who were randomly assigned to therapeutic hypothermia or standard care, therapeutic hypothermia reduced mortality (25 versus 34 percent; RR 0.75, 95% CI 0.64-0.88; RD -9 percent, 95% CI -13 to -4 percent; NNT 11, 95% CI 8-25) [35]. A 2012 meta-analysis, with seven trials including 1214 neonates, reported similar risk reduction for death (RR 0.75, 95% CI 0.63-0.88) [42].
●Improved neurodevelopmental outcomes – Therapeutic hypothermia also reduced rates of major neurodevelopmental disability among survivors at 12 to 24 months of age (26 versus 39 percent; RR 0.67, 95% CI 0.55-0.80). Major neurodevelopmental disability was generally defined as delayed mental or motor development (ie, a score ≥2 standard deviations below the mean on standardized testing), cerebral palsy, or sensory impairment (blindness or deafness). Therapeutic hypothermia had a beneficial effect on each of these outcomes, though the differences in rates of sensory impairment were not statistically significant [35]:
•Severe developmental delay – 25 percent in the hypothermia group versus 35 percent in the control group (RR 0.74, 95% CI 0.58-0.94)
•Cerebral palsy – 23 versus 35 percent, respectively (RR 0.66, 95% CI 0.54-0.82)
•Blindness – 6 versus 10 percent (RR 0.62, 95% CI 0.38-1.01)
•Deafness – 4 versus 6 percent (RR 0.66, 95% CI 0.35-1.26)
The benefits of therapeutic hypothermia were similar for the subgroups with moderate encephalopathy at baseline and severe encephalopathy at baseline [35]. The 2012 meta-analysis had similar findings for these outcomes [42].
Data regarding the long-term safety and efficacy of therapeutic hypothermia suggest that the benefits extend into later childhood. For example, in the TOBY trial, with outcome data available for 280 of the original 325 participants, infants who received therapeutic hypothermia were more likely to survive to age six or seven years with an intelligence quotient [IQ] score of ≥85 compared with infants in the control group (52 versus 39 percent, RR 1.31, 95% CI 1.01-1.71) [44]. Similarly, the NICHD trial assessed outcomes at ages six to seven years for 190 of the original 209 trial participants and found lower rates of cognitive impairment (defined as IQ score <70) in children assigned to the hypothermia group compared with the control group (47 versus 62 percent); however, the difference did not reach statistical significance (RR 0.78, 95% CI 0.61-1.01) [43].
●Benefits similar among different methods of cooling – The trials included in the 2013 meta-analysis used different cooling methods (selective head cooling in five trials with 526 neonates and whole-body cooling in five trials with 942 neonates) [35]. In subgroup analyses, the benefits of therapeutic hypothermia were consistent regardless of the method of cooling used. Similar benefits for selective head cooling and whole-body cooling were reported in a 2012 meta-analysis [42].
●Despite benefit of cooling, adverse outcomes remain common – While these clinical trial data demonstrate a clear benefit of therapeutic hypothermia, adverse neurodevelopmental outcomes remain common despite therapeutic hypothermia [45-47]. This was illustrated in the meta-analysis described above, in which only 54 percent of all infants who were treated with hypothermia survived to age 12 to 24 months without major neurodevelopmental disability [35]. In addition, the trials in the meta-analysis excluded neonates with severe intrauterine growth restriction, mild encephalopathy, or prematurity. Additional studies on therapeutic hypothermia and other neuroprotective therapies are urgently needed for these groups of patients [18,48]. (See 'Range of outcomes' below.)
●Unclear if delayed cooling is beneficial – It is unclear if delayed administration of therapeutic hypothermia (ie, started after six hours of life) conveys similar benefits as early therapeutic hypothermia. In a randomized trial addressing this question, 168 neonates with moderate to severe HIE were randomly assigned to therapeutic hypothermia starting at 6 to 24 hours after birth and continuing for 96 hours or to normothermia (maintained at 37°C [98.6°F]) [49]. Mortality was similar in both groups (11 percent in each). The trial did not detect a significant difference in rates of moderate or severe disability at 18 to 22 months of age (13 versus 17 percent; adjusted RR 0.89, 95% CI 0.54-1.48). Based on the results of this single small trial, delayed hypothermia does not appear to convey a meaningful benefit. However, it is possible that the trial was underpowered to detect a difference.
No apparent benefit in low- and middle-income countries — Randomized trials performed in low- and middle-income countries (LMIC) have failed to demonstrate a benefit of therapeutic hypothermia and suggest that the therapy may be harmful in this setting [50-52]. The multicenter HELIX trial, performed in tertiary NICUs in India, Sri Lanka, and Bangladesh, randomly assigned 408 term infants with moderate or severe neonatal encephalopathy who were within six hours of birth to whole-body hypothermia using a servo-controlled cooling device or to usual care (control group) [50]. At 18 to 22 months, the composite outcome of death or moderate or severe disability was similar in both groups (50 versus 47 percent; RR 1.06, 95% CI 0.87-1.30), while death alone was increased for the hypothermia group (42 versus 31 percent; RR 1.35, 95% CI 1.04-1.76). Several secondary outcomes were also worse in the hypothermia group.
The negative findings of the HELIX trial contrast with findings of trials done in high-income countries. The reasons for the discrepant results are not clear, but differences in the populations studied may be a factor. Unlike infants in high-income countries, most infants in the HELIX trial were born outside of the hospital, and the quality of preadmission care could not be evaluated [50]. In addition, HELIX included a higher proportion of infants who were small for gestation age or low birth weight, had clinical seizures at enrollment, or had magnetic resonance imaging (MRI) markers suggesting a longer duration of hypoxic-ischemic injury [50,51].
For mild hypoxic-ischemic encephalopathy — Whether therapeutic hypothermia improves outcomes in infants with milder degrees of encephalopathy is unknown. Studies suggest that newborns with mild encephalopathy can also exhibit brain injury [53,54]; however, the safety and efficacy of therapeutic hypothermia for mild encephalopathy has not been established [55,56].
Neonates who are not candidates for cooling — Neonates who are not candidates for therapeutic hypothermia (eg, preterm neonates or those with mild encephalopathy) should receive supportive care measures, as described above (see 'Supportive management' above). In addition, we suggest close monitoring of core body temperature and strict avoidance of hyperthermia. Although it is unproven whether maintaining normothermic body temperature improves outcome in this setting, secondary analyses of infants in the control groups of the NICHD and CoolCap trials found a significant association between elevated temperature and adverse outcome [57-59]. These observational data do not establish causality, and additional studies are needed to determine whether maintaining normothermia improves neurologic outcomes in infants who do not receive therapeutic hypothermia. Nevertheless, since this intervention is unlikely to be harmful and has the potential to improve outcomes, we suggest maintaining normothermia in neonates not receiving therapeutic hypothermia.
Future prospects — A variety of potential neuroprotective treatments are being studied to prevent the cascade of injurious effects after hypoxia-ischemia [60-63]. As an example, erythropoietin has neuroprotective properties in animal models of hypoxic-ischemic brain injury and neonatal stroke [64]. However, a placebo-controlled randomized trial of 500 infants undergoing therapeutic hypothermia for HIE found that erythropoietin did not reduce the risk of death or neurodevelopmental impairment [65]. Infants in the trial who received erythropoietin also experienced more serious adverse events during the neonatal period; therefore, high doses of erythropoietin should not be given to infants undergoing therapeutic hypothermia for HIE. Lower-quality evidence suggests possible neuroprotective effect of monotherapy with erythropoietin (ie, without hypothermia) [66], but larger trials are needed to confirm benefit, which could apply to situations where cooling is not an option (eg, in LMIC [50]) or where cooling is not proven to be effective (eg, mild HIE).
Other investigational therapies include the following:
●Darbepoetin, a long-acting erythropoiesis-stimulating agent, as monotherapy for milder neonatal encephalopathy and in combination with therapeutic hypothermia [67].
●Allopurinol, a xanthine oxidase inhibitor, which reduces the production of oxygen radicals that damage mitochondria [67]. Small clinical studies before the era of therapeutic hypothermia suggested possible neuroprotective benefits of allopurinol [68]. A phase III randomized controlled trial of allopurinol and hypothermia is underway in neonatal encephalopathy [69].
●Melatonin has been evaluated in small pilot studies as an adjunct to therapeutic hypothermia; lower-quality evidence suggests a possible neuroprotective benefit of melatonin, but further study is needed to establish evidence of benefit [70].
●Human umbilical cord blood mononuclear cells have been evaluated in feasibility studies for moderate to severe neonatal encephalopathy [71,72]. Preliminary data show safety and feasibility of cell collection and preparation.
PROGNOSIS —
The likelihood and extent of brain damage and adverse outcomes is related to the severity of encephalopathy.
Range of outcomes — Therapeutic hypothermia reduces death and disability among infants with neonatal encephalopathy, but outcomes remain suboptimal. In a systematic review of 1214 neonates with hypoxic-ischemic encephalopathy (HIE) who were treated with therapeutic hypothermia, nearly one-half either died or had major neurodevelopmental disability at 18 months, while 40 percent had a normal neurologic outcome [42].
Among survivors, permanent neurologic sequelae of neonatal brain injury can be mild, such as learning difficulties or attention deficit hyperactivity disorder, or may be severe and disabling, such as cerebral palsy, epilepsy, visual impairment, or severe cognitive and developmental disorders. One report of 110 school-aged survivors of neonatal encephalopathy found that subnormal intelligence quotient (IQ) scores at six to seven years of age were present in more than 25 percent of children overall; an IQ score <70 among survivors with and without cerebral palsy was found in 96 and 9 percent, respectively [45].
Cerebral palsy develops in approximately 13 percent [73]. Although there is an increased risk of cerebral palsy associated with neonatal encephalopathy, it is not an inevitable consequence. In most cases of cerebral palsy or later developmental deficits, the cause is related to conditions other than prior neonatal encephalopathy. (See "Cerebral palsy: Epidemiology, etiology, and prevention".)
Epilepsy occurs more frequently in children with prior neonatal encephalopathy. Approximately 25 percent of children with neonatal seizures and neonatal encephalopathy developed post-neonatal epilepsy in one study [74]. Severe neonatal encephalopathy is a risk factor for infantile spasms [75].
Clinical predictors — Although definitions vary, a more severe degree of neonatal encephalopathy, as categorized by modified Sarnat criteria (table 2), and the presence of seizures are associated with increased risk of adverse outcome [74,76,77].
Most term infants with mild neonatal encephalopathy develop normally [78,79], although many will have evidence of brain injury on MRI [53,80]. Infants with moderate to severe encephalopathy are more likely to develop long-term neurologic morbidity [78,79,81-84]. Infants with moderate encephalopathy have a 20 to 35 percent risk of later sequelae from the insult, although those whose neurologic examinations are completely normal within one week and whose brain MRI show no evidence of injury have a good likelihood of normal outcome [82,85].
In the era before the advent of therapeutic hypothermia, infants with severe encephalopathy had a 75 percent risk of dying in the neonatal period and, among survivors, an almost universal risk of sequelae [79,82,86,87].
Neuroimaging predictors — Brain MRI and magnetic resonance spectroscopy (MRS) strongly predict long-term outcome following neonatal encephalopathy [88-91]. Typically, brain MRI is performed after the infant is rewarmed, at four to seven days of age when any diffusion-weighted imaging abnormalities can still be seen. Neonatal brain MRI and MRS are validated and well-accepted biomarkers of HIE severity, neurologic outcome [92-97], and treatment response following hypothermia [98-103]. Severe brain MRI abnormalities are usually associated with marked EEG abnormalities and poor outcome. (See 'EEG predictors' below.)
●Lesion location and pattern – Abnormal signal in the posterior limb of the internal capsule appreciated on a brain MRI obtained in the first two weeks of life has been shown to predict adverse neurologic outcome [88,104]. In term infants with neonatal encephalopathy, lesions affecting bilateral basal ganglia and thalami that are detected by MRI in the first weeks of life have been associated with more severely impaired neurologic outcomes and death, particularly when more extensive injury to the perirolandic cortex, other cortical regions, and brainstem are present [88,92,105,106]. A study of 451 infants with neonatal encephalopathy and brain MRI at a median age of five days showed severe brain injury on MRI was strongly associated with more severe outcomes; however, infants with mild/moderate MRI brain injury had cognitive, language, and motor scores at two years that were similar to those of infants with no injury [107].
Some reports suggest that a watershed pattern of brain injury (ie, involving the boundary regions of the major cerebral vascular territories) on neonatal brain MRI is associated with long-term cognitive, language, and motor deficits, even among children without major disability such as cerebral palsy [93,108-110]. However, extensive isolated watershed distribution signal abnormalities on diffusion-weighted brain MRI are not invariably associated with a poor short-term outcome, even when cystic evolution of injury occurs [111].
●Diffusion imaging – Diffusion-weighted MRI can detect the presence of acute brain injury; that is, injury that occurred within 7 to 10 days before the study. Thus, diffusion-weighted images can distinguish which infants with neonatal encephalopathy have suffered a significant brain injury within a window of time that often includes the time of delivery [89,90,112,113].
Diffusion abnormalities on MRI resolve in the second week after injury while the underlying brain injury continues to evolve [114]. As diffusion abnormalities resolve, the extent of injury may not be fully apparent on T1- and T2-weighted MRI. Brain MRI obtained during this period of diffusion pseudonormalization between days 7 and 10 should be interpreted within the context of this limitation.
●Impact of therapeutic hypothermia – In one study performed before the advent of therapeutic hypothermia, 30 percent of infants with neonatal encephalopathy demonstrated a completely normal head MRI during the newborn period, indicating a good prognosis [92]. In infants with HIE who are treated with therapeutic hypothermia (see 'Therapeutic hypothermia' above), the rates of normal brain MRI are even higher, ranging from 41 to 60 percent [98,99,115,116]. The available evidence suggests that treatment with hypothermia does not affect the value of MRI for predicting outcome after neonatal encephalopathy [98,99,101,116].
●MRI scoring systems – In a report of 117 infants (gestational age ≥36 weeks) with HIE who received therapeutic hypothermia, a deep learning model predicted adverse motor outcomes at 12 to 24 months based on three factors: putamen/globus pallidus injury on T1 MRI; gestational age; and cord pH [117]. Several MRI scoring systems are published [118,119] but are mainly used for research; none is optimal for predicting neurodevelopmental outcomes [120,121].
●Magnetic resonance spectroscopy – MRS can detect increased lactate and decreased N-acetyl aspartate (NAA) concentrations, which indicate derangements of the metabolic state of specific brain regions and portend a worse prognosis [122-125]. A prospective multicenter study using 3 Tesla MRI found that thalamic NAA concentrations measured within 14 days after birth predicted adverse neurodevelopmental outcomes at two years with a sensitivity and specificity of 100 and 97 percent, respectively [126]. Earlier reports had suggested a lower predictive value for MRS. In a 2010 meta-analysis of single-center studies, elevated Lac/NAA ratios in the thalamus or basal ganglia demonstrated a pooled sensitivity of 82 percent and specificity of 95 percent for neurodevelopmental outcome [127], and a 2013 systematic review found that MRS was not as predictive of outcome as other MRI parameters [90].
EEG predictors — Findings on EEG and amplitude-integrated EEG (aEEG) can provide useful prognostic information in neonatal encephalopathy. Therefore, most centers that treat infants with moderate to severe encephalopathy will perform EEG monitoring for at least 24 hours, or longer if electrographic seizures are seen [128]. Other cooling centers perform continuous EEG (cEEG) monitoring throughout cooling and rewarming to evaluate for the presence of subclinical seizures and assess change in EEG background over time [129].
An EEG that shows seizure activity or severe background abnormalities including burst suppression, isoelectricity, or extremely low voltage portends an increased likelihood of death or significant long-term neurologic sequelae [130-135]. Since severe brain MRI abnormalities are usually associated with marked EEG abnormalities and poor outcome, the EEG may be especially helpful as a prognostic tool in the setting of moderate MRI abnormalities [136,137].
A 2012 systematic review identified 29 observational studies that evaluated 13 prognostic tests applied to term infants (n = 1306) with HIE who had at least 18 months of follow-up [90]. When obtained within the first week after birth, the best performing prognostic tests were aEEG (pooled sensitivity and specificity, 0.93 and 0.90) and routine EEG (pooled sensitivity and specificity, 0.92 and 0.83). However, the confidence intervals for these data were wide because of small patient numbers in the included studies.
In infants with neonatal encephalopathy undergoing therapeutic hypothermia, a persistently abnormal aEEG at 48 hours of age is more predictive of adverse outcome than an abnormal aEEG at an earlier age. A 2017 meta-analysis included nine studies with individual patient data that evaluated the prognostic accuracy of aEEG at various time points (6, 24, 48, or 72 hours) after birth in term infants (n = 520) with neonatal encephalopathy treated with therapeutic hypothermia [138]. The main finding was that a persistently severe abnormal aEEG at ≥48 hours was predictive of an adverse neurodevelopmental outcome (ie, death or moderate to severe disability) at one year or more. In the pooled data, an abnormal aEEG at 48 hours had a good sensitivity (85 percent) and specificity (93 percent), while an abnormal aEEG at 6 hours had a good sensitivity (96 percent) but a poor specificity (39 percent).
In a subsequent multicenter study of term infants with neonatal encephalopathy undergoing therapeutic hypothermia (n = 142), infants with a severe background abnormality at any time point in the cEEG (during a mean cEEG recording time of approximately 90 hours) were significantly more likely to die or have severe neurodevelopmental impairment at age two years [139]. Similarly, in a single-center retrospective study of term and near-term infants with HIE (n = 117), infants with background abnormalities on aEEG (during a recording time of ≥84 hours) and a slower rate of aEEG recovery within 48 hours were more likely to have neurodevelopmental impairment at age two years and ages five to eight years [134].
EEG abnormalities during the period of therapeutic hypothermia are also predictive of the development of post-neonatal epilepsy. Predictors of epilepsy after acute provoked neonatal seizures include ≥3 days of seizures on EEG and severe background abnormalities [140].
Biomarkers — There are no established biomarkers for determining the extent of neonatal brain injury or predicting outcome in infants with neonatal encephalopathy. However, several studies have found that elevated levels of serum tau protein are associated with worse outcomes in HIE [141-143].
In a systematic review published in 2009, serum interleukin-1B, serum interleukin-6, cerebrospinal fluid neuron-specific enolase (NSE), and cerebrospinal fluid interleukin-1B (all measured before age 96 hours) were putative predictors of abnormal outcomes at age ≥12 months in survivors of neonatal encephalopathy [144]. In another systematic review of infants with neonatal encephalopathy published in 2023, levels of serum NSE and S100 calcium binding protein B (S100B) collected in the first 72 hours of life were associated with a worse short- and long-term prognosis [145]. However, the studies included in these reviews generally had small patient numbers and significant heterogeneity.
Another study of 180 infants enrolled in a randomized trial of erythropoietin as an adjunctive therapy to therapeutic hypothermia found that biomarkers of neuroinflammation modestly improved estimate of two-year outcomes compared with clinical data alone. These biomarkers included interleukin-6 and NSE at baseline and interleukin-8, tau, and ubiquitin carboxy-terminal hydrolase-L1 at day 4 [146].
Further studies are needed to determine if any of these biomarkers is useful for the early assessment of infants with neonatal encephalopathy.
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Neonatal encephalopathy".)
SUMMARY AND RECOMMENDATIONS
●Supportive care – Neonates with moderate and severe neonatal encephalopathy should be cared for in a neonatal intensive care unit (NICU). Key aspects of supportive care include (see 'Supportive management' above and "Perinatal asphyxia in term and late preterm infants", section on 'Supportive care based on organ system'):
•Maintain adequate ventilation; avoid hypoxemia, hyperoxia, and respiratory acidosis (see "Overview of mechanical ventilation in neonates")
•Maintain adequate perfusion; avoid hypotension or hypertension (see "Neonatal shock: Management")
•Maintain normal fluid, electrolyte, acid-base status, and euglycemia; provide adequate nutrition (see "Fluid and electrolyte therapy in newborns" and "Management and outcome of neonatal hypoglycemia")
•Control seizures (see "Treatment of neonatal seizures")
•Consult a metabolic/genetic specialist if a metabolic disorder is suspected (see "Metabolic emergencies in suspected inborn errors of metabolism: Presentation, evaluation, and management" and "Inborn errors of metabolism: Identifying the specific disorder")
●Therapeutic hypothermia – For term or late preterm infants with moderate to severe neonatal encephalopathy (table 2) who are within the first six hours after birth, we recommend therapeutic hypothermia (table 1) (Grade 1B). Therapeutic hypothermia can be provided with either head or whole-body cooling targeting a core (rectal) temperature of 33 to 35°C (91.4 to 95°F) for 72 hours. This recommendation applies to experienced centers in high-income countries and may not apply to low- or middle-income countries where the benefit is uncertain. (See 'Therapeutic hypothermia' above.)
●Noncandidates for cooling – For newborns with neonatal encephalopathy who are not candidates for therapeutic hypothermia, it is reasonable to closely monitor core body temperature, maintaining normothermia and strictly avoiding hyperthermia. However, whether maintaining normothermic body temperature improves outcome in this setting is unproven. (See 'Neonates who are not candidates for cooling' above.)
●Prognosis – Most infants with mild encephalopathy develop normally, while infants with moderate to severe encephalopathy are more likely to develop long-term neurologic morbidity. Severe brain MRI abnormalities are usually associated with marked EEG abnormalities and more severe neurodevelopmental impairment. Permanent neurologic sequelae can be mild, such as specific learning difficulties or attention deficit hyperactivity disorder, or may be more severe, such as cerebral palsy, visual impairment, and severe cognitive and developmental disorders. (See 'Prognosis' above.)
