OBJECTIVE. Single-channel amplitude-integrated electroencephalography has been shown to be predictive of neurodevelopmental outcome in term infants with hypoxic-ischemic encephalopathy. We describe the relationship of quantifiable electroencephalogram (EEG) measures, obtained using a 2-channel digital bedside EEG monitor from term newborn infants with encephalopathy and/or seizures, to cerebral injury defined qualitatively by MRI.
METHODS. Median values of minimum, mean, and maximum EEG amplitude were obtained from term-born encephalopathic infants during a 2-hour seizure-free period obtained within 72 hours of admission. Infants underwent MRI with images qualitatively scored for abnormalities of cortex, white matter, deep nuclear gray matter, and posterior limb of the internal capsule. Eighty-six infants had EEG measures related to qualitative MRI outcomes.
RESULTS. The most common diagnosis was hypoxic ischemic encephalopathy (n = 40). For all infants there was a negative relationship between EEG amplitude measures and MRI abnormality scores assessed on a scale from 4 to 15, with a higher score indicating more abnormalities. This relationship was strongest for the minimum amplitude measures in both hemispheres; that is, for every unit increase in score there was a mean drop of 0.41 μv for the left cerebral hemisphere, with 35% of variance explained. This relationship persisted on sub-group analyses for infants with hypoxic-ischemic encephalopathy, infants with other diagnoses and infants monitored after the first 24 hours of life. Using an MRI abnormality score cutoff of 8 or worse for cerebral injury in infants with hypoxic-ischemic encephalopathy, a minimum amplitude of 4 μV showed a higher specificity (80%: left hemisphere), whereas a minimum amplitude of 6 μV showed a higher sensitivity (92%: left hemisphere).
CONCLUSIONS. Bedside EEG measures in term-born encephalopathic infants are related to the severity of cerebral injury as defined by qualitative MRI. A minimum amplitude of <4 μV appears useful in predicting outcome.
Although bedside electroencephalogram (EEG) monitors have been increasingly used over the last 2 decades, their role in the management of infants with newborn encephalopathy remains to be fully established. Neonatal encephalopathy in term-born infants is an important clinical problem,1 with severely affected infants being at a substantial risk of dying or developing cerebral palsy.2,3 The use of early conventional EEG in this group of infants has been shown to be predictive of long-term neurologic outcome.4,5 Over the last 2 decades, the single-channel amplitude-integrated EEG (aEEG) has been increasingly used in newborn encephalopathic infants. It is convenient and can monitor the background EEG activity over long periods. In term-born encephalopathic infants, it has also been shown to be predictive of neurologic outcome,6,7 particularly in combination with an early neurologic examination.8 aEEG findings have been used for selecting infants for neuroprotective intervention studies, such as hypothermia.9,10
Newer, 2-channel bedside EEG monitors provide both raw EEG and quantitative measures of EEG trace, in addition to the aEEG trace amplitude, from the cerebral cortex of each hemisphere.11 Such bilateral hemispheric measures may increase the sensitivity for the detection of unilateral cerebral injury. The aim of the present study was to relate quantitative measures of EEG amplitude, obtained from encephalopathic term newborn infants, to the severity of cerebral injury defined by qualitative MRI.
Between November 2001 and November 2004, infants admitted to the tertiary NICUs at the Royal Children's and Royal Women's Hospitals, Melbourne, with newborn encephalopathy and/or seizures who had EEG monitoring as part of their clinical care were recruited into the study. All of the infants underwent a period of ≥4 hours of continuous bedside EEG monitoring with the 2-channel bedside EEG monitor, the BRM2 (BrainZ Instruments, Auckland, New Zealand). In addition all of the infants underwent MRI at a time based on the clinician's guidance as to the stability of the infant.
Diagnoses were made on clinical grounds by the attending physician. The diagnosis of hypoxic-ischemic encephalopathy (HIE) was based on history and examination findings, such as the presence of perinatal distress, including diminished fetal movements, cardiotocograph abnormalities, fetal acidosis, meconium staining of liquor, low Apgar scores, and the need for neonatal resuscitation. The stage of encephalopathy was graded for all of the infants with HIE using a modified Sarnat clinical classification12 with maximum score within 72 hours of admission being noted. Infants with stage 1 HIE demonstrated hyperalertness, hyper-reflexia, tachycardia, jitteriness, and dilated pupils. Infants with stage 2 HIE were lethargic, with a bradycardia, hypotonia, constricted pupils, weak suck, poor Moro response, and, usually, seizures. Infants with stage 3 HIE demonstrated stupor, flaccidity, hypotonia, hyper-reflexia, and absent suck, gag, and Moro reflexes. Infants with other diagnoses were deemed to be encephalopathic on a clinical basis, and severity was graded in relation to level of consciousness. A combination of features, including abnormality of tone, poor suck, abnormal gag or Moro response, central respiratory depression with a need for mechanical ventilation, and seizures were recorded.
Bedside EEG Monitoring
The BRM2 is a 2-channel bedside monitor that uses 5 electrodes. These were placed in the C3, P3, C4, P4, and reference positions of the 10–20 system.13 The 2 channels of EEG result from a potential difference between the C3 and P3 and between the C4 and P4 electrodes, respectively. The monitor also has the capacity to provide a single “cross-hemisphere” channel view, the potential difference between the P3 and P4 as used in single-channel aEEG studies.6,7 In practice, the distance between the central and parietal electrodes was the same for all of the infants at 2.5 cm. The monitor allows continuous recording and stores the information to digital memory.
The “raw” EEG traces were manually reviewed for seizure activity. An optimum 2-hour seizure-free trace with an electrode impedance of <20 kΩ was analyzed at the earliest feasible period of recording by a single observer (D.K.S.) blinded to MRI outcomes. The raw EEG traces were analyzed offline using Labview Chartanalyzer software (modified from National Instruments Corporation, Austin, TX), which provides median values for amplitude measures over any chosen time period. In infants with a normal background, the period analyzed incorporated an approximately equal number of sleep and wake cycles. Median values for minimum, mean, and maximum amplitudes were obtained for the left and right cerebral hemispheres separately.
The aEEG was also classified into 3 groups of background patterns: normal (minimum amplitude >4 μV, maximum amplitude >9 μV), moderately abnormal ([MA] minimum amplitude <4 μV, maximum amplitude >9 μV), and severely abnormal ([SA] minimum amplitude <4 μV, maximum amplitude <9 μV). The limits of 4 and 9 μV were chosen instead of 5 and 10 μV used by other groups,7 because the distance of 2.5 cm between the central and parietal electrodes is smaller than the interelectrode distance used with the single biparietal channel devices. Hence, all of the amplitude measures obtained would be lower than with the single biparietal device channel. Locating electrodes closer to each other may potentially provide more focal information from the cerebral cortex. However, the lower amplitudes are thought to result from lead currents flowing more within skin tissue, decreasing the signal sensitivity to the brain tissue and increasing the noise, hence resulting in lower amplitudes.14
Qualitative Magnetic Resonance Assessment
Ninety-three infants underwent MRI. The magnetic resonance (MR) protocol used a standard transmit receive head coil (GE Healthcare, Waukesha, WI) including T1-weighted images (repetition time [TR]/echo time [TE]: 420/12; field of view [FOV]: 18 × 14 cm; matrix: 256 × 224; slice thickness: 3 mm; echo train length: 2; averages: 3), T2-weighted images (TR/TE: 3540/16; FOV: 20 × 18 cm; matrix: 256 × 224; slice thickness: 3 mm; echo train length: 21; averages: 3), and diffusion-weighted sequences (TR/TE: 10 000/105; FOV: 25 × 19 cm; matrix: 192 × 128; slice thickness: 4 mm; averages: 2; b: 1; 3 directions [x,y,z]).
The MRIs were analyzed qualitatively by a single rater (T.E.I.), blinded to the EEG recording, using a scoring system in which the cortex, the white matter signal (including ventricular size), deep nuclear gray matter (DNGM), and posterior limbs of the internal capsule (PLIC) were graded, and a cumulative MR abnormality score (MRAS) from 4 to 15 was obtained for each cerebral hemisphere (Table 1 and Fig 1).
For the assignment of the score on MRI of the white matter, cortex, and DNGM, a score of 1 was awarded if the tissue was normal; a score of 2 was awarded for ≤2 small focal regions of abnormality; a score of 3 was awarded for involving up to half of the region within the hemisphere; and a score of 4 if more than half the region within the hemisphere was involved. Mild ventricular dilatation would be awarded a score of 2, moderate dilatation a score of 3, and severe dilatation a score of 4 for the white matter domain. If myelination of the PLIC was present and normal, a score of 1 was awarded; if myelination was present but impaired, a score of 2 was awarded; and if absent, a score of 3 was awarded.
An MRAS cutoff at ≥8 was chosen, because infants with this score demonstrated the presence of cerebral abnormalities of a severe nature in 1 region or moderate nature in >1 region. This cutoff was used to define the groups of infants with normal-mild versus moderate-severe cerebral injury. One of every 10 MRIs was rescored, and the intraobserver variability was 5%.
Statistical Analysis of Results
Data analysis was conducted using the SPSS version 11.5 (SPSS Inc, Chicago, IL) statistics software package. Linear regression was used to relate the amplitude outcomes to the MRAS.
Between November 2001 and November 2004, 235 term-born infants were cared for with a diagnosis of HIE and/or seizures between the 2 neonatal units. Ninety five (40%) of these infants had bedside monitoring with the BRM2 over this period. A retrospective review of a clinical cohort of infants who had had bedside EEG monitoring was conducted. Monitoring was not available continuously, and, hence, some infants could not be monitored. For the remainder, the decision to monitor was made by the attending neonatologist based on personal clinician preference, severity of encephalopathy, and suspicion of seizures. There was a small difference in the infants selected for monitoring in relation to use of anticonvulsants (monitored: 66%; not monitored: 70%) and mortality (monitored: 19%; not monitored: 15%).
Clinical characteristics of the infants are detailed in Table 2. Seventy-four percent of the infants were outborn, and there were more male infants. Two thirds of the infants were treated with anticonvulsants for clinical and/or electroencephalographic seizures. The infants received phenytoin, phenobarbitone, diazepam, clonazepam, or midazolam. Nineteen percent of the infants died before discharge. HIE was the most common diagnosis; the clinical characteristics of these infants are shown in Table 3. The median age at EEG monitoring for all of the infants was 2.2 days (range: 0–14 days).
Of the 95 infants who had bedside EEG monitoring, 93 had MRI, and 2 died before MRI was undertaken. Seven infants had continuous or frequent seizures (>3 seizures an hour) throughout the entire period during which they were being monitored. These 7 traces were, therefore, not analyzed, because the background EEG amplitude measures may have been altered because of the seizure activity and not reflective of true background activity. Thus, 86 infants had EEG amplitude measures related to qualitative MRI analysis. Only 13 of the infants had EEG monitoring conducted in the first 12 hours of life. MRI was conducted at a median time of 4 days after the bedside EEG monitoring (range: 2 days before monitoring to 55 days after monitoring; n = 93).
Quantitative Amplitude in Relation to Severity of MRI
On linear regression, there was a significant negative relationship between all of the EEG amplitudes in both hemispheres and MR scores (Table 4). For every unit increase in MRAS there was a mean drop of 0.41 μv in minimum amplitude (95% confidence interval [CI]: −0.29 to −0.53 μv; P < .001), 35% of variance explained, for the left cerebral hemisphere (Fig 2) and 0.36 μv (95% CI: −0.23 to −0.49 μv) for the right cerebral hemisphere.
A subgroup analysis of the patients who died showed that all of the infants except 1 had minimum amplitudes <5 μV (minimum amplitudes [μV]; median [range]: left: 1.8 (0.8–7.1); right: 2.0 (0.8–7.5); n = 16). The exception was an infant who had Opitz syndrome who died of causes not related to encephalopathy.
Qualitative Background Pattern in Relation to MRI
On analysis of variance, infants with MA and SA background patterns had significantly greater MRAS than infants with a normal background EEG pattern (normal versus MA versus SA; n [50 vs 17 vs 19]; mean [95% CI]; 6.7 [6.0 to 7.4] vs 9.2 [7.7 to 10.7] vs 10.9 [9.3 to 12.5]; P < .001; Fig 3).
Relationship of Timing of aEEG and MRI
For all of the infants, there was a broad scatter of minimum amplitudes in relation to the age at which bedside monitoring was conducted (Fig 4). Similarly, there was no relationship between MRAS for all of the infants and timing of EEG monitoring (Fig 4) or the time interval between EEG monitoring and MRI acquisition. Even at later ages, there was a marked scatter of MRAS values.
Pattern of MR Injury
Fifteen infants had a pattern of predominantly severe DNGM injury (DNGM score 3 or 4 with white matter abnormality score 1 or 2), and 18 had a pattern of predominantly severe white matter injury (white matter abnormality score 3 or 4 with DNGM 1 or 2). On analysis of variance, increasing severity of DNGM injury resulted in decreasing minimum amplitudes (score 1 vs 2 vs 3 vs 4; [n] 32 vs 19 vs 25 vs 10; mean [SD] μV; 5.5 (2.0) vs 4.7 (2.2) vs 4.0 (2.1) vs 2.3 (1.5); P < .001, left hemisphere).
For infants with isolated severe DNGM injury, there was a negative relationship between minimum amplitude and MRAS. For every unit increase in MRAS, there was a mean decrease of 0.72 μV in minimum amplitude (95% CI: −1.2 to −0.2 μV; P = .009; 39% variance explained). Five of the 86 infants (6%) had markedly asymmetric brain injury (a difference in MRAS ≥3 between the 2 hemispheres), which was reflected in the recording channel.
Infants Diagnosed With HIE
A subgroup analysis was conducted for the 40 infants who were diagnosed to have HIE (Table 3). A similar but stronger relationship was observed between all of the amplitude measures and MRAS. For every unit increase in MRAS, there was a mean drop of 0.41 μv in minimum amplitude (95% CI: −0.26 to −0.56 μv; P < .001, 44% of variance explained for the left cerebral hemisphere) and 0.40 μv (95% CI −0.26 to −0.53 μv) for the right cerebral hemisphere (P < .001; 49% of variance explained; Table 4).
Infants with stage 2 or 3 HIE (Sarnat classification12 modified) were more likely to have a lower minimum amplitude than those with stage 1 (Fig 5). For all of the infants who died who had an analyzable trace, the median value for the minimum amplitude was 1.8 μv (range: 0.8–6.3 μv; n = 18).
Infants With Diagnoses Other Than HIE
On carrying out a subgroup analysis for the 46 infants with encephalopathy because of causes other than HIE, a direct relationship persisted between minimum amplitude and MRAS. For every unit increase in MRAS, there was a drop in minimum amplitude by 0.31 μv (left hemisphere; 95% CI: −0.5 to −0.1 μV; P = .005), with 17% variance explained. For this group of infants, the median age at EEG monitoring was 3 days (range: 0–14 days), and the median age at MRI was 10 days (range: 1–63 days).
Infants Monitored After the First 24 Hours of Life
Fifty eight of the 86 infants were monitored after the first 24 hours of life. The median age of monitoring was 2.9 days (range: 1–14 days). Subgroup analysis for this group of infants showed a direct relationship between MRAS and minimum amplitude. For every unit increase in MRAS there was a drop in minimum amplitude by 0.35 μv (left hemisphere; 95% CI: −0.51 to −0.20 μV); P < .001), with 27% variance explained.
Infants With Seizures
The 7 infants whose traces could not be analyzed because of frequent seizures had a median MRAS close to the SA end of the spectrum, with only 1 infant with normal MRAS (4 for each hemisphere), and the rest of the infants with scores ranging from 12 to 15 for each hemisphere. The infant with the normal MRAS had a relatively short period of monitoring with no 60-minute seizure-free period. Between seizures, the aEEG background was discontinuous (MA).
Effect of Anticonvulsants
On adding the use of anticonvulsants as a predictor in the linear regression model, no significant effect was observed (regression coefficient: −0.05; 95% CI: −0.92 to 0.81; P = .90 for the right cerebral hemisphere; regression coefficient: 0.04; 95% CI: −0.80 to 0.88; P = .93 for the left cerebral hemisphere), and, hence, there was no need to adjust for an anticonvulsant effect. When the analysis was confined for infants who had received anticonvulsants (n = 63), the relationship between minimum amplitude and MRAS persisted. For every unit increase in MRAS, there was a mean drop of 0.36 μv in minimum amplitude (95% CI: −0.21 to −0.51 μv; P < .001), with 30% of variance explained for the left cerebral hemisphere and 0.31 μv (95% CI: −0.13 to −0.48 μv) for the right cerebral hemisphere (P = .001), with 19% of variance explained.
For 36 of the 63 infants who had received anticonvulsants (57%), accurate records of the time interval between the last anticonvulsant dose and EEG monitoring were available. This ranged from 40 minutes to 135 hours. On linear regression between the minimum amplitude (left) and the time interval between the last dose of anticonvulsant and EEG monitoring, there was a marked scatter of values with no significant relationship between the time of the last anticonvulsant dose and the minimum amplitude (regression coefficient: 0.02; 95% CI: −0.01 to 0.05; P = .18).
For 7 patients, amplitude measures immediately before and 30 minutes after anticonvulsant administration were obtained. Three infants had received 5 mg/kg maintenance doses of phenobarbitone, 1 had received 10 mg/kg of phenobarbitone, 2 had received loading doses of 20 mg/kg of phenobarbitone, and 1 had received 15 mg/kg of phenytoin. The amplitude measures from the 2 hemispheres were analyzed together. A related-samples t test showed that the amplitude measures decreased significantly in the minimum amplitude (mean difference: −0.86 μV; 95% CI: −1.31 to − 0.42 μV; P = .001) and mean amplitude (mean difference: −1.06 μV; 95% CI: −1.77 to − 0.36 μV; P = .007) but not the maximum amplitude (mean difference: −1.12 μV; 95% CI: −2.77 to 0.52 μV; P = .16).
Diagnostic Accuracy of EEG for More Severe Cerebral Injury in Infants With HIE
For the 40 infants with HIE, using an MRAS cutoff at ≥8 as abnormal, a minimum EEG amplitude of ≤4 μV provided good specificity (eg, sensitivity: 72%; specificity: 80%; positive predictive value [PPV]: 86%; negative predictive value [NPV]: 63%, left hemisphere; sensitivity: 75%; specificity: 66%; PPV: 78%; NPV: 63%, right hemisphere), whereas a minimum amplitude of ≤6 μV showed a higher sensitivity (eg, sensitivity: 92%; specificity: 33%; PPV: 70%; NPV: 71%, left hemisphere; sensitivity: 100%; specificity: 13%; PPV: 65%; NPV: 100%, right hemisphere; Table 5). The value of 4 μV was chosen on the basis of a receiver operating curve plot. There were no substantial differences in diagnostic accuracy between hemispheres (data for right hemisphere not shown).
Diagnostic Accuracy of EEG for More Severe Cerebral Injury in Infants Monitored After the First 24 Hours of Life
Using a similar process for the 58 infants who were monitored after the first 24 hours of life, similar results were obtained (minimum EEG amplitude of 4 μV; sensitivity: 50%; specificity: 72%; PPV: 62%; NPV: 64%, left hemisphere; sensitivity: 42%; specificity: 69%; PPV: 52%; NPV: 59%, right hemisphere; minimum EEG amplitude of 6 μV; sensitivity: 85%; specificity: 41%; PPV: 54%; NPV: 76%, left hemisphere; sensitivity: 81%; specificity: 44%; PPV: 54%; NPV: 74%, right hemisphere).
This study provides evidence that quantifiable amplitude EEG measures from term infants who present with encephalopathy and/or seizures are significantly related to the extent of cerebral injury on qualitative MRI. Minimum amplitude shows the strongest relationship to this outcome. Similar relationships were obtained for the subgroups of infants who were diagnosed to have HIE and in infants who had already been treated with anticonvulsants. In the subgroup with HIE, differing amplitude cutoffs altered sensitivity and specificity for more severe cerebral injury, with a minimum amplitude EEG of 4 μV providing a higher specificity, whereas a minimum amplitude EEG of 6 μV had a greater sensitivity. These findings are consistent with those of al Naqeeb et al,7 who used a 5-μV cutoff for the minimum amplitude with a single-channel device. Lower amplitudes would be expected for the central-parietal channel used in the present study as compared with the single biparietal channel because of a smaller interelectrode distance. The current study was conducted using the Brainz BRM2, a 2-channel device with the electrodes placed in the C3, P3, C4, P4, and reference positions. In principle, these results should be applicable to any other system using similar electrode placement, that is, 2.5 cm.
The neurologic status on clinical examination has been the cornerstone for prognosticating the outcome for infants with newborn encephalopathy.12,15,16 However, more recently, electroencephalographic and neuroimaging studies have entered the NICU as useful aids in evaluating the extent of cerebral injury in the term infant. Severity of abnormality of conventional EEG5 and aEEG6,7 background in infants with newborn encephalopathy has been shown to be prognostic of poor neurodevelopmental outcome, particularly when used early. There is evidence that prognosis is also related to the speed at which the aEEG trace normalizes.17 This electrical background abnormality may parallel cerebral metabolism and secondary energy failure, although interface mechanisms remain to be elucidated.18
aEEG has been used as part of the enrollment criteria for neuroprotection trials, such as the hypothermia studies.9,10 In this context, the predictive value of the aEEG is thought to lie in the fact that it demonstrates abnormalities in the first few hours of life, at a time when neuroprotective therapies may have the most benefit. It was also demonstrated in the recent head cooling trial that the aEEG trace on entry into the study delineated which infants may benefit most from this therapy; selective head cooling was more beneficial to infants with MA aEEG traces in the first 5.5 hours of life rather than infants with more SA aEEGs.10 Thus, aEEG patterns assisted in recruitment and predicting potential benefit from selective hypothermia. These findings require further confirmation.
Previous aEEG studies have concentrated on infants suffering from encephalopathy because of hypoxia-ischemia, and within this group, the optimum predictive value was shown to be in the first few hours of life.6,19 Our work extends these findings to include infants with encephalopathy from a variety of causes and shows that EEG amplitude measures are depressed in the presence of cerebral injury from causes other than hypoxia-ischemia. Additional studies would assist in the understanding of the evolution of the aEEG findings to delineate more regionally and mechanistically specific patterns of cerebral insults in the newborn brain.
The presence of electrographic and clinical seizures in newborn encephalopathy has been associated with a poor prognosis,16,20 but the extent to which they are causative in affecting adverse outcome remains unclear.21 The treatment of seizures in newborn encephalopathy is known to affect the EEG background,22,23 and some preliminary work suggests that anticonvulsants may have a more prolonged suppressive effect on the aEEG traces of infants with poor background patterns compared with infants who have better backgrounds.23 Although this study was able to demonstrate a decrease in EEG measures immediately after the administration of anticonvulsants, we found little effect of anticonvulsants on the relationship between aEEG amplitudes and cerebral injury as assessed by MRAS. When analyses were confined to infants who had received anticonvulsants, the relationship between minimum amplitude and MRAS persisted. This suggests that, in our study, the nature of the underlying neuropathology was more dominant than the effect of anticonvulsant therapy and is important in reassuring clinicians in evaluating the aEEG trace in infants who have received anticonvulsant therapy.
In our study, the median time at which EEG monitoring commenced was considerably outside the first 6 hours when the single-channel aEEG has been shown to be most predictive of long-term neurologic outcome.6 Most of the infants studied were outborn and referred to tertiary centers from a wide geographic area. Hence, this was a select group of sick infants reflected in the relatively high number of infants who had received anticonvulsants for clinical seizures, who required mechanical ventilation, and who died. Although the EEG measures were obtained from the earliest analyzable trace, they represent a “snapshot.” It is clear that aEEG background patterns evolve over time after a cerebral insult,17 but even if there is delay in monitoring, quantitative amplitude seems useful as a prognostic aid. Our data confirm that even when used after the first 24 hours of life, bedside EEG has good prognostic use.
Our outcome measure in this study was that of qualitative MRI analysis. MRI reporting is subject to bias, and there is no universally standardized approach for qualitative reporting. We attempted to address intraobserver variability by repeating every tenth assessment. Also, a cumulative MRAS covering all the regions of the brain was used instead of a qualitative score based solely on the cerebral cortex. Although the EEG signal at the scalp represents the sum of the excitatory and inhibitory postsynaptic potentials from neuronal cells in the cortex, it results from complex interactions and feedback loops among the cortex, DNGM, and other regions of the brain.13 It is indeed reassuring that the prognostically important pattern of isolated DNGM injury24 is equally well reflected using surface aEEG recordings.
This study demonstrates that aEEG is prognostically useful in relation to MR outcomes up until 3 days after presentation to a NICU in a broad group of encephalopathic infants and in the presence of anticonvulsant treatment. A minimum amplitude of 4 μV gives a specificity of 75%, which may allow the bedside monitor to be used as a tool for prognosticating in conjunction with examination and neuroimaging findings, whereas a higher minimum amplitude of 6 μV would allow the tool to be used as a screening test for diagnostic purposes or for consideration for neuroprotective interventions, with its higher sensitivity. Longer-term neurodevelopmental outcome measures will allow further validation of the bedside EEG monitor in this group of infants.
We acknowledge BrainZ Instruments, New Zealand, for their support of this study.
We thank the MRI department at the Royal Children's Hospital, Melbourne, for their continued support.
We would also like to thank Dr Jeffrey Neil for his input.
- Accepted February 1, 2006.
- Address correspondence to Terrie E. Inder, MD, Department of Pediatrics, St Louis Children's Hospital, Washington University, One Children's Place, St Louis, MO 63108. E-mail:
Financial Disclosure: Dr Shah is the recipient of a PhD scholarship from BrainZ New Zealand. Ms Lavery and Ms Wong were research nurses partially funded by BrainZ New Zealand.
- ↵Volpe JJ. Hypoxic-ischemic encephalopathy: clinical aspects. In: Volpe JJ, ed. Neurology of the Newborn 4th ed. Philadelphia, PA: WB Saunders; 2001:331– 394
- ↵Dixon G, Badawi N, Kurinczuk JJ, et al. Early developmental outcomes after newborn encephalopathy. Pediatrics.2002;109 :26– 33
- ↵Vannucci RC, Perlman JM. Interventions for perinatal hypoxic-ischemic encephalopathy. Pediatrics.1997;100 :1004– 1014
- ↵Biagioni E, Mercuri E, Rutherford M, et al. Combined use of electroencephalogram and magnetic resonance imaging in full-term neonates with acute encephalopathy. Pediatrics.2001;107 :461– 468
- ↵Toet MC, Hellstrom-Westas L, Groenendaal F, Eken P, de Vries LS. Amplitude integrated EEG 3 and 6 hours after birth in full term neonates with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal. Ed.1999;81 :F19– F23
- ↵al Naqeeb N, Edwards AD, Cowan FM, Azzopardi D. Assessment of neonatal encephalopathy by amplitude-integrated electroencephalography. Pediatrics.1999;103 :1263– 1271
- ↵Shalak LF, Laptook AR, Velaphi SC, Perlman JM. Amplitude-integrated electroencephalography coupled with an early neurologic examination enhances prediction of term infants at risk for persistent encephalopathy. Pediatrics.2003;111 ;351– 7
- ↵Azzopardi D, Robertson NJ, Cowan FM, Rutherford MA, Rampling M, Edwards AD. Pilot study of treatment with whole body hypothermia for neonatal encephalopathy. Pediatrics.2000;106 :684– 694
- ↵Inder TE, Buckland L, Williams CE, et al. Lowered electroencephalographic spectral edge frequency predicts the presence of cerebral white matter injury in premature infants. Pediatrics.2003;111 :27– 33
- ↵Fisch BJ. Fisch and Sphlmann's EEG Primer, Basic Principles of Digital and Analog EEG 3rd revised and enlarged ed. New York, NY: Elsevier; 1999
- ↵Rutherford M, Pennock J, Schwieso J, Cowan F, Dubowitz L. Hypoxic-ischaemic encephalopathy: early and late magnetic resonance imaging findings in relation to outcome. Arch Dis Child Fetal Neonatal Ed.1996;75 :F145– F151
- Copyright © 2006 by the American Academy of Pediatrics