A new electroencephalogram classification with reduced recording time in asphyxiated term infants

Brain & Development 36 (2014) 372–379 www.elsevier.com/locate/braindev

Original article

A new electroencephalogram classification with reduced recording time in asphyxiated term infants Toru Kato a,⇑, Takeshi Tsuji a, Fumio Hayakawa a, Tetsuo Kubota b, Hiroyuki Kidokoro c, Jun Natsume c, Kazuyoshi Watanabe d, Akihisa Okumura e a

Department of Pediatrics, Okazaki City Hospital, Okazaki, Japan b Department of Pediatrics, Anjo Kosei Hospital, Anjo, Japan c Department of Pediatrics, Nagoya University Graduate School of Medicine, Nagoya, Japan d Faculty of Health and Medical Science, Aichi Shukutoku University, Nagoya, Japan e Department of Pediatrics and Adolescent Medicine, Juntendo University School of Medicine, Tokyo, Japan Received 6 April 2013; received in revised form 22 May 2013; accepted 10 June 2013

Abstract Background and objectives: Conventional electroencephalogram (cEEG) is a reliable predictor of outcome in term infants with hypoxic ischemic encephalopathy (HIE). Early therapeutic hypothermia initiated within 6 h after birth is a beneficial treatment in these infants. However, a classification system with reduced cEEG recording time to determine early intervention has not been reported. The aim of this study is to propose a new classification of depression on cEEG with reduced recording time in infants with HIE and to examine the correlation between the classification and short-term outcome. Patients and methods: We retrospectively investigated 20 term infants with HIE in whom cEEG was performed within 12 h after birth, and deaths or outcomes at 18 months of age were assessed. We determined grades 0–3 EEG depression in each 10-min epoch based on the most common EEG patterns of each 20 s epoch defined by our criteria. Results: Eighteen infants could be assessed by depression grade. The Spearman’s rank correlation coefficient Rs between the maximum depression grade in 10-min epochs and three-grade outcomes was 0.68 (P = 0.002), and that between the minimum one and outcomes was 0.66 (P = 0.003). The area under the receiver operating characteristic curve of the maximum and minimum depression grades for predicting abnormal outcome were 0.885 and 0.869, respectively. Conclusions: We demonstrated a new cEEG depression classification with a recording time of at least 10 min in term infants with HIE and a good correlation with short-term outcome. Ó 2013 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved. Keywords: Classification; Electroencephalogram; Hypoxic ischemic encephalopathy; Prediction; Term infant

1. Introduction Hypoxic ischemic encephalopathy (HIE) is a major cause of death and neurological impairment in asphyxiated neonates [1]. Randomized clinical trials have demonstrated the benefit of therapeutic hypothermia commenced within 6 h after birth for reducing death or disability at 18 months ⇑ Corresponding author. Address: Department of Pediatrics, Okazaki City Hospital, 3-1 Goshoai, Koryuji-cho, Okazaki, Aichi 444-8553, Japan. Tel.: +81 564 21 8111; fax: +81 564 25 5531. E-mail address: [email protected] (T. Kato).

of age [2–4]. In 2010, the American Heart Association, European Resuscitation Council, and International Liaison Committee on Resuscitation recommended therapeutic hypothermia in term or near-term infants with moderate to severe HIE [5]. Conventional electroencephalogram (cEEG) is a highly reliable predictor of outcome in term infants with HIE [6–9]. However, the previous cEEG classification criteria require a recording time of 45–120 min because assessments of sleep–wake cycles or cEEG patterns during all sleep stages are needed for most of these criteria.

0387-7604/$ - see front matter Ó 2013 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.braindev.2013.06.007

T. Kato et al. / Brain & Development 36 (2014) 372–379

A classification system with reduced cEEG recording time in infants with HIE has not been reported. On the other hand, early initiation of appropriate interventions such as therapeutic hypothermia is important for a good prognosis in these infants. Therefore, clinicians must decide between a long-duration assessment of cEEG and earlier intervention of therapeutic hypothermia. Short-duration and precise evaluation of brain function in infants with HIE will aid in early determination and provision of appropriate interventions for a good prognosis. Our hypothesis based on our experience is that it is possible to evaluate cEEG with reduced recording time of at least 10 min within 12 h after birth in term infants with HIE to predict short-term outcome at 18 months of age. The aim of this study was to develop criteria for classifying cEEG depression with reduced recording time in term infants with HIE and to clarify the correlation between the classification and short-term outcome. 2. Methods 2.1. Patients We retrospectively investigated 20 consecutive term infants with HIE admitted to the neonatal intensive care units of Okazaki City Hospital (N = 8), Anjo Kosei Hospital (N = 9), Toyota Memorial Hospital (N = 1), Japanese Red Cross Nagoya Daiichi Hospital (N = 1), and Nagoya University Hospital (N = 1) between April 2004 and March 2010. These hospitals provide tertiarylevel care for newborns. The infants fulfilled all of the following criteria: (1) P37 weeks gestation; (2) Apgar score 65 at 5 min after birth, pH 6 7.0 in umbilical-cord blood or a blood sample immediately after birth or base deficit P 16 mmol/l in the same sample, or assisted ventilation for at least 5 min after birth; (3) clinical signs of moderate or severe encephalopathy as defined by Shankaran et al. [2]; (4) no congenital anomalies or inborn errors of metabolism; (5) cEEG performed within 12 h after birth; and (6) assessment of death or neurological outcome at 18 months of age. We collected the clinical data of all subjects from the medical charts. 2.2. cEEG recordings cEEG was recorded polygraphically at the bedside for at least 10 min using a bipolar montage with at least eight surface Ag/AgCl cup electrodes (AF3–C3, C3–O1, AF4–C4, C4–O2, AF3–T3, T3–O1, AF4–T4, and T4–O2) according to the international 10–20 system. This was combined with an electrocardiogram and respiratory movement assessment by the NicoletOne Monitor nICU (Natus Medical, San Carlos, CA, USA) or Nihon Kohden EEG Neurofax (Nihon Kohden, Tokyo, Japan). The low-cutoff filter was fixed at 0.5 Hz, and

373

the raw cEEG was displayed at 3 cm/s and 100 lV/ 10 mm. Impedance was <20 kX. The preparation time of set-up for recording was within 10–15 min in our hospitals. Written informed consent for the cEEG recordings was obtained from the parents of all patients. Approval of the institutional ethics committee at Okazaki City Hospital was obtained to conduct this study. 2.3. cEEG assessments Representative cEEG patterns in the 20 s epochs are shown in Fig. 1. We evaluated cEEG patterns in each 20 s epoch throughout recording using criteria based on our previous report, with minor modifications, to determine the depression grade on cEEG [6,10]. The cEEG patterns were classified as follows: P, poor activity of 5–20 lV or inactivity of 0–5 lV; B, a burst–suppression pattern consisting of abnormal burst activity of >50 lV for <4 s interrupted by attenuated activity of 0–20 lV; D, discontinuous pattern consisting of >4 s physiological burst activity interrupted by attenuated activity of 0–20 lV; dA, depressed-alternating pattern consisting of physiological high-voltage slow bursts of 50–150 lV separated by low-voltage activity of 20– 50 lV, but the duration of the bursts was <3 s or the duration of low-voltage activity >8 s; L, continuous low-voltage irregular activity of 20–50 lV; A, alternating pattern consisting of physiological high-voltage slow bursts of 50–150 lV for 4–8 s separated by low-voltage activity of 20–50 lV for <8 s; M, continuous mediumvoltage activity of 30–100 lV; H, continuous highvoltage slow activity of 50–150 lV; S, seizure activity consisting of stereotyped, repetitive, and rhythmic activity lasting >10 s; N, noise or artifacts; U, unclassified pattern consisting of patterns other than the above. Minor modifications included addition of the new dA pattern and a more detailed definition of patterns B and D. If inter-hemispheric asymmetry or asynchrony was seen, we selected the former pattern, which was more depressed. The expert investigator (T. Kato) who alone evaluated the cEEG patterns was blinded to the clinical information. The inter- and intra-rater agreement on these patterns was assessed in 15 representative samples of 20 s epochs consisting of all patterns except for S, N, and U by three authors (T. Kato, T. Tsuji, and T. Kubota) who are experts at evaluating neonatal cEEG using j statistics. Among 15 samples, disagreements between two raters were seen in two to three samples (j = 0.82). The intra-rater agreement (T. Kato) was perfect in the 15 samples (j = 1). Thereafter, we extracted the most common pattern in the 20 s epochs in each 10-min record as the representative 10-min epoch pattern (Fig. 2). We evaluated all consecutive 10-min epochs whether in wakefulness or in sleep. If pattern S was seen in at least one 20 s epoch, then the representative pattern in this 10-min epoch

374

T. Kato et al. / Brain & Development 36 (2014) 372–379

Fig. 1. Representative electroencephalogram (EEG) patterns in a 20 s epoch. Representative EEG patterns of a 20 s epoch are shown. (A) Poor activity. (B) Burst–suppression pattern. (C) Discontinuous pattern. (D) Depressed-alternating pattern. (E) Low-voltage irregular activity. (F) Alternating pattern. (G) Medium-voltage activity. (H) Continuous high-voltage slow activity. EEG patterns are defined in the text. EEGs are arrayed in the order of AF3–C3, C3–O1, AF4–C4, C4–O2, AF3–T3, T3–O1, AF4–T4, and T4–O2. Calibration shows 1 s and 100 lV.

Fig. 2. Extraction of the representative electroencephalogram (EEG) pattern in each 10-min epoch. Scatter plots are shown as dots representing each 20-s epoch in the recorded EEG sample. Patterns of EEG are defined in the text: P, poor activity; B, burst–suppression pattern; D, discontinuous pattern; dA, depressed-alternating pattern; L, low-voltage irregular activity; A, alternating pattern; M, mediumvoltage activity; H, continuous high-voltage slow activity; S, seizure activity; U, unclassified pattern; and N, noise. Arrows shows the most common patterns in each 10-min epoch, which are extracted as the representative pattern in the respective epoch.

was considered unable to be evaluated because seizure activity would affect interictal activities such as postictal suppression. If the most common pattern in a 20 s epoch was U or N, then the representative pattern

in this 10-min epoch was also considered unable to be evaluated. If the most common patterns in 20 s epochs were the same in number between two different patterns, we selected the pattern that was more depressed as the representative pattern for that 10-min epoch. Finally, we determined the depression grade in each 10-min epoch of all records as follows: grade 0, the most common pattern was H, M, or A; grade 1, the most common pattern was L, dA, or D; grade 2, the most common pattern was B; grade 3, the most common pattern was P; and unable to grade, the most common pattern was N or U, or the grade could not be evaluated because of S. We assessed the maximum and the minimum cEEG depression grade in each 10-min epoch separately in each record. A list of EEG pattern definitions and the depression grades are shown in Table 1. 2.4. Outcome measurements Early infantile death caused by HIE was determined from medical charts. Neurodevelopmental outcome was assessed by neonatologists or child neurologists at 18 months of age. Both motor and cognitive impairments were evaluated based on a complete neurological examination and developmental assessments. We assessed the motor disability grade at 18 months of age as levels 1–5 based on the Gross Motor Function

T. Kato et al. / Brain & Development 36 (2014) 372–379

375

Table 1 Definitions of the electroencephalogram (EEG) patterns in each 20-s epoch and depression grade. Pattern

Amplitude, duration or characteristics of EEG activity

Depression gradea

P B D dA

Amplitude of 0–20 lV Abnormal burst activity of >50 lV for <4 s interrupted by attenuated activity of 0–20 lV Physiological burst activity for >4 s interrupted by attenuated activity of 0–20 lV Physiological high-voltage slow bursts of 50–150 lV for <3 s, or interruption of low voltage activity of 20–50 lV for >8 s Continuous activity of 20–50 lV Physiological high-voltage slow bursts of 50–150 lV for 4–8 s separated by low-voltage activity of 20–50 lV for <8 s Continuous activity of 30–100 V Continuous activity of 50–150 V Seizure activity consisting of stereotyped, repetitive, and rhythmic activity lasting >10 s Noise or artifacts Unclassified pattern

3 2 1

L A M H S N U

0

Unable to grade

Patterns of EEG are defined in the text. If inter-hemispheric asymmetry or asynchrony was seen, the upper pattern was selected. a The most common pattern in a 20-s epoch during a 10-min epoch was converted to a depression grade.

Classification System (GMFCS) [11]. We assessed the speech ability at 18 months of age by hearing from parents and examinations, and also assessed having epilepsy as sequelae of HIE. We defined disability as the presence of motor disability of level 2 or higher on the GMFCS, speech delay with no vocabulary, or epilepsy at 18 months of age. We graded the overall outcome as grade 0, normal; grade 1, disability as defined above; and grade 2, death. We also defined abnormal outcome as grades 1 or 2 for the overall outcome. 2.5. Statistical analysis Data were analyzed with the R software package (ver. 2.8.1; http://www.R-project.org) or IBM SPSS statistics version 20 (IBM, Armonk, NY, USA). Spearman’s rank correlation coefficient Rs calculated with a two-sided test was used to evaluate the correlation between cEEG grade and outcome. The Rs value was judged as follows: >0.7, strongly positive; 0.4–0.7, fairly positive; 0.2–0.4, weakly positive; and 0–0.2, negligible correlation. Receiver operating characteristic curves were used to assess the predictive value of cEEG grade on abnormal outcome. P-values <0.05 were considered significant. 3. Results

Table 2 Subject characteristics (N = 20). Demographic data Gestational age, wks, mean (SD) Birth weight, g, mean (SD) Male gender (N) Apgar score at 5 min, median (range) Inborn neonates (N) Placental abruption (N) Umbilical cord prolapse (N) Feto-maternal hemorrhage (N) Fetal heart rate monitoring (N) Variable deceleration (N) Late deceleration (N) Prolonged bradycardia (N) Cord or blood gas pH, mean (SD) Cord or blood gas base deficit, mmol/l, mean (SD) Mechanical ventilation (N) Therapeutic hypothermia (N) Clinical signs of encephalopathy Moderate (N) Severe (N) EEG initiated time after birth, h, mean (SD)

39.6 (1.0) 3014 (419) 11 4 (0–7) 9 5 1 1 19 9 1 8 7.06 (0.22) 17.9 (8.6) 17 14 11 9 5.4 (3.4)

and severe in nine infants. Placental pathology was examined in seven patients. Among them, chorioamnionitis was seen in three patients, and placental abruption was seen in one patient. There were no differences of therapeutic strategies and outcomes among the hospitals.

3.1. Patient characteristics 3.2. cEEG recording The subject characteristics are shown in Table 2. Subjects had a mean (SD) gestational age of 39.6 (1.0) weeks and birth weight of 3014 (419) g. Fetal heart rate monitoring before delivery was performed in 19 patients, and severe abnormality was seen in 18 patients. Therapeutic hypothermia initiated within 6 h after birth using selective head cooling was performed in 14 (70%) infants. Clinical signs of encephalopathy were moderate in 11

The start of cEEG recording time was a mean (SD, range) of 5.4 (3.4, 1.8–11) h after birth. Only one infant underwent therapeutic hypothermia with phenobarbital and midazolam during cEEG recording; all other infants received no sedative or anticonvulsive agents or therapeutic hypothermia during cEEG recording. cEEG was recorded before the initiation of therapeutic

376

T. Kato et al. / Brain & Development 36 (2014) 372–379

hypothermia in 13 infant. The mean (SD) duration of cEEG recording was 32 (15) min. 3.3. cEEG depression grade Only 17 of 1920 (0.9%) 20 s epochs were evaluated as U. Among 20 subjects, two infants had an S cEEG pattern in every 10-min epoch of the entire record; thus, we were unable to grade cEEG depression. The depression grade in each 10-min epoch was evaluated in another 18 infants, who were finally the subjects for assessment of correlation between the depression grade and outcome. Among them, 15 showed the same cEEG depression grade in every 10-min epoch throughout the record. Another three infants showed different cEEG depression grades in the 10-min epochs during recording. Two infants showed grades 0 and 1 in different 10-min epochs, and one infant showed grades 2 and 3. Finally, in 18 infants in whom the grade of cEEG depression could be evaluated, the maximum grade was 0 in two infants, 1 in four, 2 in two, and 3 in 10; the minimum grade was 0 in four, 1 in two, 2 in three, and 3 in nine (Table 3). 3.4. Outcomes

infants, level 2 in three, level 3 in 0, level 4 in one, and level 5 in four. Speech delay was seen in nine infants. Epilepsy was observed in three infants until 18 months of age. 3.5. Correlation between the cEEG depression grade and outcome The correlation between the cEEG depression grade and outcome is shown in Table 4. The Spearman’s rank correlation coefficient Rs between the maximum grade of cEEG depression and the overall outcome grade at 18 months of age was 0.68 (P = 0.002), and that between the maximum cEEG depression grade and the GMFCS level at 18 months of age was 0.82 (P < 0.001). The Rs between the minimum cEEG depression grade and the overall outcome grade was 0.66 (P = 0.003), and that between the minimum cEEG depression grade and the GMFCS level was 0.77 (P < 0.001). These correlations were fairly to strongly positive and statistically significant. No significant correlation was observed between the maximum or minimum cEEG depression grade and speech delay or epilepsy. 3.6. Predictive value of cEEG grade for abnormal outcome

Among 18 infants in whom the grade of cEEG depression could be evaluated, 13 had abnormal outcomes including three deaths and 10 disabilities. In the 15 surviving infants, the GMFCS was level 1 in seven

The predictive value of cEEG grade for abnormal outcome is shown in Table 5. A cutoff point of grade 0 vs. 1, 2, and 3 in the maximum cEEG depression grade

Table 3 The details of electroencephalogram (EEG) patterns in a 10-min epoch and EEG depression grade for individual patients. Patient no.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Representative EEG pattern in each 10-min epoch 1st

2nd

P P P P P P P P P P B B L L dA M M M S S

P P P P P P P P B B B dA L dA A H M S S

3rd

4th

P

P

P N

N

EEG depression grade 5th

6th

P

N

M

P B B B dA L M M H

L A

M

S S

S S

S

7th

dA L

M

Maximum

Minimum

3 3 3 3 3 3 3 3 3 3 2 2 1 1 1 1 0 0 UG UG

3 3 3 3 3 3 3 3 3 2 2 2 1 1 0 0 0 0 UG UG

Abbreviations: GMFCS, gross motor function classification system; P, poor activity; B, burst–suppression pattern; dA, depressed-alternating pattern; L, low-voltage irregular activity; A, alternating pattern; M, medium-voltage activity; H, continuous high-voltage slow activity; S, seizure activity; N, noise; UG, unable to grade.

T. Kato et al. / Brain & Development 36 (2014) 372–379

377

Table 4 Correlation between electroencephalogram (EEG) depression grade and outcome at 18 months of age. Maximum EEG depression grade 0

1

2

3

Overall outcome, N = 18 Grade 0 (normal) Grade 1 (disability) Grade 2 (death)

2 0 0

1 3 0

2 0 0

0 7 3

GMFCS, N = 15 Level 1 Level 2 Level 3 Level 4 Level 5

2 0 0 0 0

3 1 0 0 0

2 0 0 0 0

0 2 0 1 4

Speech delay, N = 15 No Yes

2 0

1 3

2 0

1 6

Epilepsy, N = 15 No Yes

2 0

3 1

2 0

5 2

Minimum EEG depression grade Rs 0.68

0.82

0.45

0.19

P

0

1

2

3

0.002 3 1 0

0 2 0

2 1 0

0 6 3

4 0 0 0 0

1 1 0 0 0

2 0 0 0 1

0 2 0 1 3

3 1

0 2

2 1

1 5

4 0

1 1

3 0

4 2

<0.001

0.09

0.51

Rs

P

0.66

0.003

0.77

<0.001

0.40

0.14

0.26

0.34

Abbreviations: Rs, Spearman’s rank correlation coefficient; GMFCS, gross motor function classification system.

Table 5 Predictive value of electroencephalogram (EEG) grade for abnormal outcome at 18 months of age. Cutoff point

Sensitivity

Specificity

Maximum EEG depression grade 0 vs. 1, 2, 3 1 0.4 0, 1 vs. 2, 3 0.77 0.6 0, 1, 2 vs. 3 0.77 1

PPV

NPV

0.81 0.83 1

1 0.5 0.63

AUC

0.885 Minimum EEG depression grade 0 vs. 1, 2, 3 0.92 0.6 0, 1 vs. 2, 3 0.77 0.6 0, 1, 2 vs. 3 0.69 1

0.86 0.83 1

0.75 0.5 0.56 0.869

Abbreviations: PPV, positive predictive value; NPV, negative predictive value; AUC, area under the receiver operating characteristic curve.

showed sensitivity of 1.0 and specificity of 0.4, and that of grades 0, 1, and 2 vs. 3 in the maximum cEEG depression grade showed sensitivity of 0.77 and specificity of 1.0. The area under the receiver operating characteristic curve (AUC) for the maximum cEEG depression grade was 0.885 (Fig. 3). The cutoff point for grade 0 vs. 1, 2, and 3 in the minimum cEEG depression grade showed sensitivity of 0.92 and specificity of 0.6, and that of grades 0, 1, and 2 vs. 3 in the cEEG minimum depression grade showed sensitivity of 0.69 and specificity of 1.0. The AUC of the minimum cEEG depression grade was 0.869 (Fig. 3). 4. Discussion The current findings demonstrate two important points. First, we propose a new cEEG depression classi-

Fig. 3. Receiver operating characteristic curve of electroencephalogram (EEG) depression grade to predict outcome. A solid line shows the maximum EEG depression grade, and a dotted line shows the minimum EEG depression grade. The AUC for the maximum EEG depression grade is 0.885, and that of the minimum EEG depression grade is 0.869.

fication with reduced recording time of at least 10 min in infants with HIE. Second, both the maximum and minimum cEEG depression grades based on this classification system showed good correlations with short-term outcomes. We were able to evaluate the cEEG depression grade in each 10-min epoch based on the most common cEEG patterns of 20 s epochs. The criteria for the cEEG patterns for the 20 s epochs were modified from our previous reports [6,10]. To apply our previous criteria for classifying cEEG depression, we needed at least a

378

T. Kato et al. / Brain & Development 36 (2014) 372–379

40-min recording duration because assessments of all cEEG patterns at each sleep stage are required for the criteria, and the physiological duration of the sleep– wake cycle is usually 40–90 min in term infants. In the new classification system, the cEEG depression grades in 10-min epochs were sometimes different within the same record. In fact, three of 18 infants showed different grades during recording. However, in these infants, both the maximum and minimum grade of cEEG depression showed similarly good correlation with short term outcomes. Therefore, we consider that reduced recording time of 10 min is a feasible method for evaluating cEEG depression in infants with HIE. We could not fully clarify whether the maximum or minimum cEEG depression grade was the better predictor of outcome because of the small sample size. However, we consider that the maximum grade was a better predictor than minimum grade because the AUC and sensitivity were slightly greater for the maximum grade than for the minimum grade. Many reports have demonstrated the high predictive value of cEEG [6–9]. In most of these studies, cEEG recording commenced more than 10 h after birth, except in the study by Murray et al. [7], in which cEEG recording was initiated a mean of 5 h after birth. In the present study, cEEG recording was initiated at a mean of 5.4 h after birth. Clarifying the significance of early cEEG findings, particularly within 6 h after birth, is important to determine the need for therapeutic hypothermia in infants with HIE. Another significant finding of the present study was the ability to obtain the cEEG recording without sedatives, antiepileptic agents, or therapeutic hypothermia in 19 of 20 subjects because cEEG recording was started before administering antiepileptic or sedative agents and therapeutic hypothermia as a rule in our hospitals. No report has evaluated cEEG without anticonvulsive or sedative agents or therapeutic hypothermia. It is important to evaluate cEEG without these agents or therapeutic hypothermia because they can depress cEEG activity. Amplitude-integrated electroencephalogram (aEEG) is a convenient tool for continuous brain-function monitoring that provides a compressed view of variations in EEG amplitude using one or two channels [12]. Many studies have demonstrated the prognostic value of aEEG in term infants with HIE in association with depressed amplitude [13–15]. However, some reports have cast doubt on the predictive value of early aEEG within 6 h after birth [16,17]. We did not evaluate aEEG in the present study. Further study will be needed to assess whether cEEG or aEEG is a more accurate predictor in infants with HIE. In the present study, we defined the HIE infants according to the criteria of Shankaran et al. [2] with some modification. These are different criteria from those of acute intrapartum event sufficient to cause cerebral palsy by the American College of Obstetricians and

Gynecologists’ Task Force on Neonatal Encephalopathy and Cerebral Palsy, American College of Obstetricians and Gynecologists, and American Academy of Pediatrics in 2003 [18]. Although there is no universally accepted definition of HIE in infants [19], we selected the subjects who had a possibility of developing an abnormal outcome caused by HIE, and would benefit from interventions such as therapeutic hypothermia. The present study had several limitations. First, the number of subjects was small because we wanted to strictly evaluate infants with HIE who received cEEG within 12 h after birth. Fourteen infants were excluded due to the lack of cEEG recordings during this period. Further prospective research with a larger number of infants with HIE undergoing early cEEG recording is necessary. Second, the duration of cEEG recording was insufficient to fully evaluate the sleep–wake cycle. However, we consider that we demonstrated some validity for using reduced recording time. Third, the effect of therapeutic hypothermia might modify the neurological outcomes. It is difficult to take account of these therapeutic effects exactly, but 10 of 14 patients underwent therapeutic hypothermia had abnormal outcomes in this study. Therefore, we considered that the effect of the therapeutic hypothermia for the results of this study would be small. In conclusion, we developed new criteria for classifying cEEG depression with reduced recording time of at least 10 min in term infants with HIE and showed a good correlation between the classification and shortterm outcome. Short-duration and precise evaluation of brain function in infants with HIE will aid in early determination and provision of appropriate interventions for a good prognosis.

Acknowledgements The authors thank the following contributors who provided the subjects’ clinical data or clinical information of eligible patients: Dr. Seiji Hayashi, (Okazaki City Hospital); Dr. Tatsuya Fukasawa, Dr. Kaname Matsusawa (Anjo Kosei Hospital); Dr. Seiko Itomi, Dr. Ayako Yasuda (Japanese Red Cross Nagoya Daiichi Hospital); Dr. Hikaru Yamamoto (Toyota Memorial Hospital); Dr. Eiko Kato (Tosei General Hospital); Dr. Hidehiko Fujimaki, Dr. Tomoya Takeuchi, Dr. Tomohide Nakata, Dr. Hiroyuki Yamamoto, and Dr. Masahiro Hayakawa (Nagoya University Hospital). References [1] Pierrat V, Haouari N, Liska A, Thomas D, Subtil D, Truffert P. Prevalence, causes, and outcome at 2 years of age of newborn encephalopathy: population based study. Arch Dis Child Fetal Neonatal Ed 2005;90:F257–61.

T. Kato et al. / Brain & Development 36 (2014) 372–379 [2] Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE, McDonald SA, Donovan EF, et al. Whole-body hypothermia for neonates with hypoxic-ischemic encephalopathy. N Engl J Med 2005;353:1574–84. [3] Gluckman PD, Wyatt JS, Azzopardi D, Ballard R, Edwards AD, Ferriero DM, et al. Selective head cooling with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised trial. Lancet 2005;365:663–70. [4] Azzopardi DV, Strohm B, Edwards AD, Dyet L, Halliday HL, Juszczak E, et al. Moderate hypothermia to treat perinatal asphyxial encephalopathy. N Engl J Med 2009;361:1349–58. [5] Perlman JM, Wyllie J, Kattwinkel J, Atkins DL, Chameides L, Goldsmith JP, et al. Part 11: neonatal resuscitation: 2010 international consensus on cardiopulmonary resuscitation and emergency cardiovascular care science with treatment recommendations. Circulation 2010;122:S516–38. [6] Watanabe K, Miyazaki S, Hara K, Hakamada S. Behavioral state cycles, background EEGs and prognosis of newborns with perinatal hypoxia. Electroencephalogr Clin Neurophysiol 1980;49:618–25. [7] Murray DM, Boylan GB, Ryan CA, Connolly S. Early EEG findings in hypoxic-ischemic encephalopathy predict outcomes at 2 years. Pediatrics 2009;124:e459–67. [8] Hamelin S, Delnard N, Cneude F, Debillon T, Vercueil L. Influence of hypothermia on the prognostic value of early EEG in full-term neonates with hypoxic ischemic encephalopathy. Neurophysiol Clin 2011;41:19–27. [9] Nash KB, Bonifacio SL, Glass HC, Sullivan JE, Barkovich AJ, Ferriero DM, et al. Video-EEG monitoring in newborns with hypoxic-ischemic encephalopathy treated with hypothermia. Neurology 2011;76:556–62. [10] Watanabe K. The neonatal electroencephalogram and sleep-cycle patterns. In: Eyre JA, editor. The neurophysiological examination of the newborn infant. London: Mac Keith Press; 1992. p. 11–47. [11] Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

379

function in children with cerebral palsy. Dev Med Child Neurol 1997;39:214–23. Hellstro¨m-Westas L, Rose´n I. Continuous brain-function monitoring: state of the art in clinical practice. Semin Fetal Neonatal Med 2006;11:503–11. 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. ter Horst HJ, Sommer C, Bergman KA, Fock JM, van Weerden TW, Bos AF. Prognostic significance of amplitude-integrated EEG during the first 72 hours after birth in severely asphyxiated neonates. Pediatr Res 2004;55:1026–33. van Rooij LG, Toet MC, Osredkar D, van Huffelen AC, Groenendaal F, de Vries LS. Recovery of amplitude integrated electroencephalographic background patterns within 24 hours of perinatal asphyxia. Arch Dis Child Fetal Neonatal Ed 2005;90:F245–51. Sarkar S, Barks JD, Donn SM. Should amplitude-integrated electroencephalography be used to identify infants suitable for hypothermic neuroprotection? J Perinatol 2008;28:117–22. Shankaran S, Pappas A, McDonald SA, Laptook AR, Bara R, Ehrenkranz RA, et al. Predictive value of an early amplitude integrated electroencephalogram and neurologic examination. Pediatrics 2011;128:e112–20. American College of Obstetricians and Gynecologists. Task force on neonatal encephalopathy and cerebral palsy, American College of Obstetricians and Gynecologists, American Academy of Pediatrics. Neonatal encephalopathy and cerebral palsy: defining the pathogenesis and pathophysiology. Washington: American College of Obstetricians and Gynecologists; 2003. Kurinczuk JJ, White-Koning M, Badawi N. Epidemiology of neonatal encephalopathy and hypoxic-ischaemic encephalopathy. Early Hum Dev 2010;86:329–38.