Clinical Examination

Last Full Review: ILCOR 2023

The brain is prone to hypoxic injury during cardiac arrest and following the return of spontaneous circulation (ROSC). Therefore, some patients will develop global cerebral edema followed by herniation and brain death within 24 hours, while others will remain comatose and are at risk for severe neurological injury. Prognostication has been used to help families and physicians make decisions to limit or withdraw life support when unfavorable neurological outcomes are expected. A recent International Liaison Committee on Resuscitation (ILCOR) systematic review (Berg et al. 2023) considered the use of functional and structural modalities to aid clinicians in predicting a good neurological outcome for infants and children with the return of circulation (ROC), including ROSC or mechanical circulation, after resuscitation from in-hospital or out-of-hospital cardiac arrest from any cause. Testing modalities evaluated in the review included clinical examination, serum brain injury biomarkers or markers of inflammation and poor perfusion, electrophysiological signals and neuroimaging.

This review relates to the use of clinical examination for prediction of favorable neurodevelopmental outcome. Of note, all evaluated tests were used in combination with other tests by clinicians in the included studies.

 

Evidence Summary

A 2023 systematic review and Consensus on Science with Treatment Recommendation (CoSTR) (Scholefield et al. 2023b; Berg et al. 2023) by ILCOR evaluated the use of clinical examination for the prediction of survival with good neurological outcome after ROC (spontaneous or mechanical) following pediatric cardiac arrest. Newborn infants and patients in hypoxic coma from causes without cardiac arrest (e.g., respiratory arrest, toxidromes, drowning, hanging) were excluded except when a subpopulation of cardiac arrest patients could be evaluated separately. Clinical examination included pupillary response using a light or automated pupillometry, level of coma (Glasgow Coma Scale [GCS] score or Full Outline of UnResponsiveness score) and brainstem reflexes. Good neurological outcome was defined as a Pediatric Cerebral Performance Category score of 1, 2 or 3, or a Vineland Adaptive Behavioral Scales Second Edition score ≥ 70. Good neurological outcome prediction for infants and children after cardiac arrest was considered imprecise with a false-positive rate above 30%, although there is no universal consensus on the acceptable limits for imprecision.

For the assessment of pupil reactivity, eight studies with 402 combined patients were included (Scholefield et al. 2022b; Berg et al. 2023). Assessment occurred at 1 hour, 6 hours to 12 hours, 24 hours and 72 hours post-resuscitation. Most studies (6/8) showed a sensitivity of greater than 82% at all assessment time points, and a false-positive rate that ranged between 3.2% to 67%. The false-positive rate was lower with testing times within 12 hours of ROC and increased at 24 hours to 72 hours, while corresponding sensitivity for predicting good neurological outcome was 100% at 48 hours to 72 hours following ROC.

Three studies evaluated use of the GCS motor score or total GCS (Scholefield et al. 2023b; Berg et al. 2023). Only one study evaluated each test using total GCS or GCS motor score cut off, or at each testing time point. At 72 hours, the sensitivity of the motor response was 100% and the false-positive rate was 23%. Brainstem reflexes—including evoked response to pain, gag reflex and cough— were assessed in two studies with 118 patients. Presence of a pain response at 6 hours to 12 hours had a 100% predictive sensitivity and a false-positive rate of 67% (Brooks and Parks 2018, 324).

Although the predictive accuracy of tests was evaluated individually, a good practice statement by ILCOR recommends that no single test should be used in isolation for prediction of good neurological outcome (Scholefield et al. 2023b; Berg et al. 2023). The ILCOR treatment recommendation stemming from this CoSTR includes a suggestion for the use of pupillary light reflex within 12 hours after ROC for predicting good neurological outcome in children after cardiac arrest. A recommendation for or against using total GCS score, GCS motor score, motor response to any stimulus, or the use of brainstem tests after ROC, could not be made for predicting good neurological outcome in children after cardiac arrest (Scholefield et al. 2023b; Berg et al. 2023).

 

Insights and Implications

The Red Cross guidelines are informed by the ILCOR systematic review and CoSTR (Scholefield et al. 2023b; Berg et al. 2023). Because the evidence for the use of total GCS score, GCS motor score, motor response to any stimulus, and brainstem tests after ROC was each from a single study with a small sample size for each test and time point—a recommendation could not be made by ILCOR for these tests. Overall, a high risk of bias was observed based on heterogeneity across studies, small study and subject numbers, and variation in test assessment, performance and outcome measurement. Limited evidence for pupillary response suggests that the highest specificity for prediction of good neurological outcome is within 12 hours of ROC following cardiac arrest. Because meta-analysis could be performed, visual assessment of forest plots was used to assess test performance. The review authors also note that, although sedatives may be a confounding factor for clinical examination testing for predicting outcome, no studies reported assessment of confounding influence of medication or excluded the presence of residual sedation at the time of clinical examination testing (Scholefield et al. 2023b; Berg et al. 2023).

Biomarkers

Last Full Review: ILCOR 2023

The brain is prone to hypoxic injury during cardiac arrest and following the return of spontaneous circulation (ROSC). Therefore, some patients will develop global cerebral edema followed by herniation and brain death within 24 hours, while others will remain comatose and are at risk for severe neurological injury. Prognostication has been used to help families and physicians make decisions to limit or withdraw life support when unfavorable neurological outcomes are expected. A recent International Liaison Committee on Resuscitation (ILCOR) systematic review (Berg et al. 2023) considered the use of functional and structural modalities to aid clinicians in predicting a good neurological outcome for infants and children with return of circulation (ROC), including ROSC or mechanical circulation, after resuscitation from in-hospital or out-of-hospital cardiac arrest from any cause. Testing modalities evaluated in the review included clinical examination, serum brain injury biomarkers or markers of inflammation and poor perfusion, electrophysiological signals and neuroimaging.

This review relates to the use of serum brain injury biomarkers or markers of inflammation and poor perfusion for prediction of favorable neurodevelopmental outcome. Of note, all evaluated tests were used in combination with other tests by clinicians in the included studies.

 

Evidence Summary

A 2023 systematic review and Consensus on Science with Treatment Recommendations (CoSTR) (Scholefield et al. 2023a; Berg et al. 2023) by ILCOR evaluated the use of lactate, lactate clearance, acid-base (pH) levels and neuronal biomarkers for the prediction of survival with good neurological or neurodevelopmental outcome after ROC (spontaneous or mechanical) following pediatric cardiac arrest. Newborn infants and patients in hypoxic coma from causes without cardiac arrest (e.g., respiratory arrest, toxidromes, drowning, hanging) were excluded from analysis except when a subpopulation of cardiac arrest patients could be evaluated separately. Good neurological outcome was defined as a Pediatric Cerebral Performance Category score of 1, 2 or 3, or a Vineland Adaptive Behavioral Scale Second edition score ≥ 70.

Good neurological outcome prediction for infants and children after ROC was considered imprecise with a false-positive rate above 30%, although there is no universal consensus on the acceptable limits for imprecision in prediction for infants and children after cardiac arrest.

For the assessment of plasma lactate, five studies were included in the ILCOR CoSTR (Scholefield et al. 2023a; Berg et al. 2023). Due to heterogeneity and confounding, meta-analyses could not be performed. Three studies (López-Herce et al. 2014, 607; Moler et al. 2017, 318; Moler et al. 2015, 1898) reported a false-positive rate of less than 7% for a lactate less than 2 millimoles per liter at under 1 hour and at 6 hours to 12 hours, but with sensitivities only ranging between 16% and 28%. A lactate less than 2 millimoles per liter at 24 hours to 48 hours was more sensitive, but the false-positive rate was much higher. For a lactate level less than 5 millimoles per liter at 24 hours, sensitivity was reported at 89%, with a false-positive rate of 17%. Clearance of lactate to normal over 48 hours was highly sensitive but had a false-positive rate of 77% (Scholefield et al. 2023a; Berg et al. 2023).

Four studies evaluated pH levels above 7.0, 7.3 and less than 7.5 at intervals during resuscitation and up to 24 hours after ROC. Reported measurements of pH after resuscitation or within an hour after ROC had a wide range of sensitivities for predicting good neurological outcome, as did a pH greater than 7.0. False-positive rates for good neurological outcomes were above 80% for nearly all pH thresholds (Scholefield et al. 2023a; Berg et al. 2023).

Various brain injury biomarkers have been studied in adults to determine if a normal or low level of the biomarker is predictive of good outcome after cardiac arrest. A single study (Fink et al. 2014, 664) with 43 children reported values for neuron specific enolase (NSE), S100 calcium-binding protein B (S100B) and myelin basic protein. For an S100B threshold level of 0.128 nanograms (ng) per milliliter (ml) at 24 hours; an NSE threshold level of 53.1 ng per ml at 24 ng per ml and 76.7 ng per ml at 48 hours; and a myelin basic protein threshold level of 5.83 ng per ml at 24 hours at 24 hours and 5.43 ng per ml at 48 hours—sensitivity for predicting good neurodevelopmental outcome was 100%, but the false-positive rate ranged from 62% to 100%. Lowering the threshold values for all three biomarkers reduced the false-positive rate to less than 6% for predicting good neurological outcome, but also reduced sensitivity between 6% and 29% (Scholefield et al. 2023a; Berg et al. 2023).

The ILCOR treatment recommendations (Scholefield et al. 2022a; Berg et al. 2023) stemming from this CoSTR begin with a good practice statement that no single test should be used in isolation for prediction of good neurological outcome. A weak recommendation suggests using a normal plasma lactate value (less than 2 millimoles per liter) up to 12 hours following ROC for predicting good neurological outcome of children after cardiac arrest. A weak recommendation suggests against using pH following ROC for predicting good neurological outcome of children after cardiac arrest.

 

Insights and Implications

The Red Cross guidelines are informed by the ILCOR systematic review and CoSTR (Scholefield et al. 2023a; Berg et al. 2023). Evidence in the ILCOR review was insufficient to make recommendations for or against the use of time-to-lactate clearance within 48 hours following ROC; or for or against the use of neuro biomarkers, such as S100B or NSE, after ROC for predicting good neurological outcome of children after cardiac arrest. The studies in this review were not designed to test prognosis and included studies from two trials of therapeutic hypothermia in children following cardiac arrest. No assessment of confounding from sedatives was reported. Lactate measurement techniques were not described in the referenced studies (i.e., whole blood, serum/plasma).

Electrophysiology

Last Full Review: ILCOR 2023

Last Update:

The brain is prone to hypoxic injury during cardiac arrest and following the return of spontaneous circulation (ROSC). Therefore, some patients will develop global cerebral edema followed by herniation and brain death within 24 hours, while others will remain comatose and are at risk for severe neurological injury. Prognostication has been used to help families and physicians make decisions to limit or withdraw life support when unfavorable neurological outcomes are expected. A recent International Liaison Committee on Resuscitation (ILCOR) systematic review (Berg et al. 2023) considered the use of functional and structural modalities to aid clinicians in predicting a good neurological outcome for infants and children with return of circulation (ROC), including ROSC or mechanical circulation, after resuscitation from in-hospital or out-of-hospital cardiac arrest from any cause. Testing modalities evaluated in the review included clinical examination, serum brain injury biomarkers or markers of inflammation and poor perfusion, electrophysiological signals and neuroimaging.

This review relates to the use of electrophysiological signals for prediction of favorable neurodevelopmental outcome. Of note, all evaluated tests were used in combination with other tests by clinicians in the included studies.

 

Evidence Summary

A 2023 systematic review and Consensus on Science with Treatment Recommendations (CoSTR) (Scholefield et al. 2023c; Berg et al. 2023) by ILCOR evaluated the use of surface bioelectrical recordings from the central nervous system such as an electroencephalogram (EEG) and evoked potentials, brainstem auditory-evoked potentials and short-latency somatosensory evoked potentials (SSEPs; testing of brain and spinal cord responses elicited by sensory stimuli) for the prediction of survival with good neurological or neurodevelopmental outcome after ROC (spontaneous or mechanical) following pediatric cardiac arrest. Newborn infants and patients in hypoxic coma from causes without cardiac arrest (e.g., respiratory arrest, toxidromes, drowning, hanging) were excluded from analysis except when a subpopulation of cardiac arrest patients could be evaluated separately.

Good neurological outcome was defined as a Pediatric Cerebral Performance Category score of 1, 2 or 3, or a Vineland Adaptive Behavioral Scale Second Edition score ≥ 70. Good neurological outcome prediction for infants and children after ROC was considered imprecise with a false-positive rate above 30%, although there is no universal consensus on the acceptable limits for imprecision in prediction for infants and children after cardiac arrest. Due to a high degree of heterogeneity and confounding, meta-analyses could not be performed.

 

Absence or Presence of Seizures

Absence or presence of seizures in children after cardiac arrest and the relationship with good neurological outcomes at hospital discharge, 6 months and 12 months, was reported by 12 studies totalling 1165 children (Scholefield et al. 2023c; Berg et al. 2023). The use of terminology and criteria from the American Clinical Electrophysiology Society (ACNS) was reported in four of the 12 studies. For the prediction of good neurological outcome, the absence of seizures beyond 24 hours after ROC had reported sensitivities ranging from 50% to 100% with a false-positive rate of 42% to 100%.

Absence of Seizures and Status Epilepticus

For absence of status epilepticus, three studies reported good neurological outcome at hospital or pediatric intensive care unit (PICU) discharge, with 2 studies using ACNS criteria for defining status epilepticus. For the prediction of good neurological outcome at PICU or hospital discharge, the absence of status epilepticus had a reported sensitivity of greater than 90%, but a false-positive rate of 81% to 91% (Scholefield et al. 2023c; Berg et al. 2023).

Absence of Myoclonic Seizures

For the prediction of good neurological outcome at hospital or PICU discharge, the absence of myoclonic seizures had a reported sensitivity of 100% in two studies, but with a high false-positive rate (Scholefield et al. 2023c; Berg et al. 2023).

Absence of N20 Somatosensory Evoked Potentials

A single study evaluated the presence or absence of negative peak at 20 milliseconds (N20) SSEPs in 12 patients for predicting good neurological outcome (Scholefield et al. 2023c; Berg et al. 2023). Sensitivity for predicting good neurological outcome was 100% at 24 hours and 48 hours, and 83% at 72 hours. Although a false-positive rate of 0% was shown at all time points, confidence intervals ranged from 0%–71%.

Normal EEG Background

A normal EEG background was defined using ACNS definitions as “normal,” “continuous and reactive,” “continuous and unreactive” and “nearly continuous.” The presence of a continuous or normal EEG background was reported in 10 studies to have a sensitivity of less than 50% at 10 of 18 testing time points for predicting good neurological outcome (Scholefield et al. 2023c; Berg et al. 2023). The false-positive rate was less than 50% in all cases.

Attenuated, Isoelectric or Flat EEG

For the prediction of good neurological outcome, the absence of an attenuated, isoelectric or flat EEG was reported in eight studies to have a sensitivity between 91% and 100%, but with a wide range of false-positive rates (0% to 83%) (Scholefield et al. 2023c; Berg et al. 2023).

Burst Suppression, Burst Attenuation or Generalized Periodic Epileptiform Discharges

The absence of burst suppression, burst attenuation or generalized periodic epileptiform discharges was reported in six studies to have a sensitivity that increased from 81% to 100% within 6 hours to 12 hours and to 100% (95% CI, 100–100) at 24 hours, 48 hours and 72 hours (Scholefield et al. 2023c; Berg et al. 2023). Although highly sensitive, a high false-positive rate of 67% to 100% was reported at all testing time points for predicting a good neurodevelopmental outcome.

Presence of a Reactive EEG

Three studies reported the presence of a reactive EEG to have a sensitivity for predicting good neurological outcome of 53% to 80% between 6 hours and 72 hours, with a false-positive rate of 7% to 27% at up to 24 hours after ROC in two studies, and 50% at 48 hours after ROC in the third study (Scholefield et al. 2023c; Berg et al. 2023).

Presence of Sleep II Architecture or Sleep Spindle EEG Features

Two studies with 123 total patients reported that the presence of sleep II architecture or sleep spindle features on an EEG had a sensitivity for predicting neurological outcome ranging between 57% and 80% at 6 hours to 12 hours and 24 hours post-ROC (Scholefield et al. 2023c; Berg et al. 2023). The false-positive rate ranged from 8.3% to 16%.

Presence of EEG Variability

The presence of EEG variability as defined by ACNS, was found in two studies to have a 60% to 80% sensitivity and an 18% to 50% false-positive rate for predicting a good neurological outcome (Scholefield et al. 2023c; Berg et al. 2023). Electroencephalogram voltage variability in a single study had a sensitivity of 75% to 100% at all time points up to 48 hours post-ROC, and a false-positive rate of 36% to 67% (Scholefield et al. 2023c; Berg et al. 2023).

Quantitative EEG Score

A quantitative 24-hour EEG background score of greater than 15 was reported in a single study to have a 94% sensitivity and 67% false-positive rate for predicting a good neurological outcome (Scholefield et al. 2023c; Berg et al. 2023).

The ILCOR treatment recommendations (Scholefield et al. 2022d; Berg et al. 2023) stemming from the CoSTR begin with a good practice statement that no single test should be used in isolation for prediction of good neurological outcome.

A weak recommendation suggests using an EEG within 6 hours to 72 hours after ROC for predicting good neurological outcome in children after cardiac arrest.

The use of EEG-specific features after ROC is suggested by ILCOR for predicting good neurological outcome, including (Scholefield et al. 2023c; Berg et al. 2023):

• Presence of sleep spindle and sleep II architecture at 12 hours to 24 hours, or
• Continuous or normal background EEG between 1 hour and 72 hours, or
• Electroencephalogram reactivity between 6 hours to 24 hours

 

A weak recommendation by ILCOR suggests against using specific EEG features after ROC to predict good neurological outcome, including (Scholefield et al. 2023c; Berg et al. 2023):

• Absence of clinical or electrographic seizures
• Absence of status epilepticus
• Absence of myoclonic epilepsy
• Absence of burst suppression, burst attenuation or generalized periodic epileptiform discharges
• Absence of attenuated, isoelectric or flat EEG

 

Insights and Implications

The Red Cross guidelines are informed by the ILCOR systematic review and CoSTR (Scholefield et al. 2023c; Berg et al. 2023). Recommendations were not made for or against a testing modality if only one study was included in a review of a test, such as for N20 SSEPs after ROC, EEG voltage variability and quantitative EEG score for predicting good neurological outcomes; additional studies are needed on these testing modalities. The review authors note that there was limited or no context of when tests were undertaken in relation to concurrent pharmacological exposure, sedation and ongoing treatments—such as therapeutic hypothermia—in patients following cardiac arrest (Scholefield et al. 2023c; Berg et al. 2023). No assessment of confounding from sedatives was reported.

Brain Imaging

Last Full Review: ILCOR 2023

The brain is prone to hypoxic injury during cardiac arrest and following the return of spontaneous circulation (ROSC). Therefore, some patients will develop global cerebral edema followed by herniation and brain death within 24 hours, while others will remain comatose and are at risk for severe neurological injury. Prognostication has been used to help families and physicians make decisions to limit or withdraw life support when unfavorable neurological outcomes are expected. A recent International Liaison Committee on Resuscitation (ILCOR) systematic review (Berg et al. 2023) considered the use of functional and structural modalities to aid clinicians in predicting a good neurological outcome for infants and children with return of circulation (ROC), including ROSC or mechanical circulation, after resuscitation from in-hospital or out-of-hospital cardiac arrest from any cause. Testing modalities evaluated in the review included clinical examination, serum brain injury biomarkers or markers of inflammation and poor perfusion, electrophysiological signals and neuroimaging.

This review relates to the use of brain imaging for prediction of favorable neurodevelopmental outcome. Of note, all evaluated tests were used in combination with other tests by clinicians in the included studies.

 

Evidence Summary

A 2023 systematic review and Consensus on Science with Treatment Recommendations (CoSTR) (Scholefield et al. 2023a; Berg et al. 2023) by ILCOR evaluated the use of brain imaging for the prediction of survival with good neurological or neurodevelopmental outcome after ROC (spontaneous or mechanical) following pediatric cardiac arrest. Newborn infants and patients in hypoxic coma from causes without cardiac arrest (e.g., respiratory arrest, toxidromes, drowning, hanging) were excluded from analysis except when a subpopulation of cardiac arrest patients could be evaluated separately. Good neurological outcome was defined as a Pediatric Cerebral Performance Category score of 1, 2 or 3, or Vineland Adaptive Behavioral Scale Second Edition score ≥ 70.

Good neurological outcome prediction for infants and children after ROC was considered imprecise with a false-positive rate above 30%, although there is no universal consensus on the acceptable limits for imprecision in prediction for infants and children after cardiac arrest. Due to a high degree of heterogeneity and confounding, meta-analyses could not be performed. Evidence was summarized for studies evaluating the use of head computed tomography (CT) imaging, brain magnetic resonance imaging (MRI) and transcranial doppler ultrasound at intervals between less than 12 hours and/or 7 days to 10 days after cardiac arrest. No studies were identified for cranial ultrasound (Scholefield et al. 2023a; Berg et al. 2023).

The CoSTR summarized evidence from three studies that used findings reported on head CT imaging performed in most cases at 24 hours or 48 hours after cardiac arrest (Scholefield et al. 2023a; Berg et al. 2023). Two of these studies assessed neurological outcome at pediatric intensive care unit or hospital discharge, and one study assessed neurological outcome at 6 months. Cytotoxic edema of the brain due to anoxia affects the gray matter preferentially and reduces the gray/white ratio towards a value of 1 on CT scan. The presence of gray-white matter differentiation on CT at 24 hours was reported to have a sensitivity ranging from 64% to 100% and a false-positive rate of 35% to 70% for predicting good neurological outcome. Absence of other CT findings—including intracranial hemorrhage, edema and lesions—were reported to have a sensitivity of 72% to 100%, but with a wide range of false-positive rates. High sensitivity was reported for absence of effacement of sulci or basal cisterns (93% to 100%), but also with a wide range of false-positive rates (Scholefield et al. 2023a; Berg et al. 2023).

A total of four studies using brain MRI imaging to predict good neurological outcomes were summarized in the ILCOR CoSTR (Scholefield et al. 2023a; Berg et al. 2023). Imaging occurred a median of three days to six days from ROC to MRI and up to 14 days. Two studies (Fink et al. 2013, 31; Fink et al. 2020, 185) reported on the absence of abnormalities in multiple regions of the brain with diffusion-weighted imaging, T1 and T2 MRI sequence at 4 days to 6 days post-ROC. Sensitivities ranged from 67% to 100% for predicting good neurological outcome but were associated with a high false-positive rate.

A single study (Kirschen et al. 2021, e719) reported the absence of any region of abnormality or restricted diffusion at a median of 4 days post-ROC to have an 88% sensitivity with a 2% false-positive rate for predicting good neurological outcome.

A single study (Yacoub et al. 2019, 103) reported apparent diffusion coefficient thresholds that were evaluated at a median of 4 days post-ROC with predicted good neurological outcome showing high sensitivity and a low false-positive rate. The same study (Yacoub et al. 2019, 103) reported that a normal MRI by quantitative reporting of absence of hypoxic ischemic injury had an 81% sensitivity and a false-positive rate of 10% for predicting good neurological outcome.

Intracranial vessel flow patterns measured on transcranial doppler ultrasound were reported in a single small study (Lin et al. 2015, 182) of cardiac arrest patients also treated with hypothermic temperature management, with the assessment of flow patterns before, during and after therapeutic hypothermia. Continuous flow velocities without reversal of diastolic flow pattern were reported to have a sensitivity of 100% and false-positive rate of 44%, while mean flow velocity within an hour of the event had a reported sensitivity of 38% and a false-positive rate of 0%.

The ILCOR treatment recommendations stemming from the CoSTR (Scholefield et al.2023a; Berg et al. 2023) begin with a good practice statement that no single test should be used in isolation for prediction of good neurological outcome. A weak recommendation suggests against using normal CT imaging at 24 hours to 48 hours from ROC for predicting good neurological outcome. A weak recommendation suggests using normal MRI between 72 hours and 2 weeks after ROC for predicting good neurological outcome.

 

Insights and Implications

The Red Cross guidelines are informed by the ILCOR systematic review and CoSTR (Scholefield et al. 2023a; Berg et al. 2023). Several considerations were noted by the authors of the CoSTR. The evidence included in the review was at a high risk of bias. There were few studies and small patient enrollment numbers. If only one study was available for a test modality, with a small patient sample size, ILCOR could not make a recommendation. A recommendation could not be made for or against the use of transcranial doppler for predicting good neurological outcome in the ILCOR CoSTR. Although a normal MRI or CT scan has a moderate-to-high sensitivity for predicting a good neurological outcome, the specificity is up to 30% and may therefore falsely predict a less favorable neurological outcome. While neuroimaging for prognostication after cardiac arrest is promising, more studies in the pediatric population are needed.