EEG monitoring; seizure; pediatric; hypoxia-ischemia; stroke; ECMO; cardiac arrest; congenital heart disease; status epilepticus
This study examines electroencephalographic (EEG) changes in children with medication resistant epilepsy treated with the ketogenic diet (KD).
Routine EEGs were obtained prior to KD initiation, then one month and three months later. Changes in EEG background slowing and frequency of interictal epileptiform discharges (IEDs) were evaluated using power spectrum analysis and manual determination of spike index. KD responders were compared to non-responders to determine if baseline or early EEG characteristics predicted treatment response (>50% seizure reduction) at three months.
Thirty-seven patients were evaluated. No differences in baseline EEG features were found between responder groups. Frequency of IEDs declined in 65% of patients as early as one month, by a median of 13.6% (IQR 2-33). Those with a ten percent or greater improvement in IED frequency at one month were greater than six times more likely to be KD responders (OR 6.5 95% CI 0.85 to 75. p=0.03). Qualitative and quantitative measures of EEG background slowing improved in the whole cohort, but did not predict responder status.
Baseline predictors of KD response remain elusive. Most patients experienced a reduction in IEDs and improvement in EEG background slowing after KD initiation. Reduction of IEDs at one month strongly predicted KD responder status at three months.
Ketogenic Diet; EEG; spike index; power spectrum analysis
Clinical neurologic signs considered predictive of adverse outcome after pediatric cardiac arrest (CA) may have a different prognostic value in the setting of therapeutic hypothermia (TH). We aimed to determine the prognostic value of motor and pupillary responses in children treated with TH after CA.
Prospective cohort study.
Pediatric ICU in tertiary care hospital.
Children treated with TH after CA.
Measurements and Main Results
Thirty-five children treated with TH after CA were prospectively enrolled. Examinations were performed by emergency medicine physicians and intensive care unit bedside nurses. Examinations were performed after resuscitation, 1 hour after achievement of hypothermia, during the last hour of hypothermia, 1 hour after achievement of normothermia, after 24 hours of normothermia, and after 72 hours of normothermia. The primary outcome was unfavorable outcome at ICU discharge, defined as a Pediatric Cerebral Performance Category (PCPC) score of 4–6 at hospital discharge. The secondary outcome was death (PCPC = 6). The associations between exam responses and unfavorable outcomes (as both PCPC 4,5,6 and PCPC 6) are presented as positive predictive values (PPV), for both all subjects and subjects not receiving paralytics. Statistical significance for these comparisons was determined using Fisher’s exact test. At all examination times and examination categories PPV is higher for the unfavorable outcome PCPC 4,5,6 than PCPC 6. By normothermia hour 24, absent motor and pupil responses were highly predictive of unfavorable outcome (PCPC 4,5,6) (PPV 100% and p<0.03 for all categories), while at earlier times the predictive value was lower.
Absent motor and pupil responses are more predictive of unfavorable outcome when defined more broadly than when defined as only death. Absent motor and pupil responses during hypothermia and soon after return of spontaneous circulation were not predictive of unfavorable outcome while absent motor and pupil responses once normothermic were predictive of unfavorable short-term outcome. Further study is needed using more robust short-term and long-term outcome measures.
Therapeutic Hypothermia; Neurological Examination; Pediatric; Hypoxic Ischemic Encephalopathy; Cardiac Arrest; Prognosis
Adverse neurodevelopmental sequelae are reported among children who undergo early cardiac surgery to repair congenital heart defects (CHD). APOE genotype has previously been determined to contribute to the prediction of these outcomes. Understanding further genetic causes for the development of poor neurobehavioral outcomes should enhance patient risk stratification and improve both prevention and treatment strategies.
We performed a prospective observational study of children who underwent cardiac surgery before six months of age; this included a neurodevelopmental evaluation between their fourth and fifth birthdays. Attention and behavioral skills were assessed through parental report utilizing the Attention Deficit-Hyperactivity Disorder-IV scale preschool edition (ADHD-IV), and Child Behavior Checklist (CBCL/1.5-5), respectively. Of the seven investigated, three neurodevelopmental phenotypes met genomic quality control criteria. Linear regression was performed to determine the effect of genome-wide genetic variation on these three neurodevelopmental measures in 316 subjects.
This genome-wide association study identified single nucleotide polymorphisms (SNPs) associated with three neurobehavioral phenotypes in the postoperative children ADHD-IV Impulsivity/Hyperactivity, CBCL/1.5-5 PDPs, and CBCL/1.5-5 Total Problems. The most predictive SNPs for each phenotype were: a LGALS8 intronic SNP, rs4659682, associated with ADHD-IV Impulsivity (P = 1.03×10−6); a PCSK5 intronic SNP, rs2261722, associated with CBCL/1.5-5 PDPs (P = 1.11×10−6); and an intergenic SNP, rs11617488, 50 kb from FGF9, associated with CBCL/1.5-5 Total Problems (P = 3.47×10−7). 10 SNPs (3 for ADHD-IV Impulsivity, 5 for CBCL/1.5-5 PDPs, and 2 for CBCL/1.5-5 Total Problems) had p<10−5.
No SNPs met genome-wide significance for our three neurobehavioral phenotypes; however, 10 SNPs reached a threshold for suggestive significance (p<10−5). Given the unique nature of this cohort, larger studies and/or replication are not possible. Studies to further investigate the mechanisms through which these newly identified genes may influence neurodevelopment dysfunction are warranted.
Continuous EEG (cEEG) monitoring is being used with increasing frequency in critically ill patients, most often to detect non-convulsive seizures. While cEEG is non-invasive and feasible in the critical care setting, it is also expensive and labor intensive, and there has been little study of its impact on clinical care. We aimed to determine prospectively the impact of cEEG on clinical management in critically ill children.
Critically ill children (non-neonates) with acute encephalopathy underwent cEEG. Study enrollment and data collection were prospective.
100 children were studied. EEG monitoring led to specific clinical management changes in 59 children. These included initiating or escalating anti-seizure medications in 43 due to seizure detection, demonstrating that a specific event (subtle movement or vital sign change) was not a seizure in 21, or obtaining urgent neuroimaging that led to a clinical change in 3. In the remaining 41 children, cEEG ruled out the presence of non-convulsive seizures but did not lead to a specific change in clinical management.
EEG monitoring led to changes in clinical management in the majority of patients, suggesting it may have an important role in management of critically ill children. Further study is needed to determine whether the management changes elicited by cEEG improve outcome.
Seizure; Status epilepticus; Pediatric; Critically Ill; Electroencephalogram; EEG monitoring
Neonatal seizures are often refractory to treatment with initial antiseizure medications. Consequently, clinicians turn to alternatives such as levetiracetam, despite the lack of published data regarding its safety, tolerability, or efficacy in the neonatal population. We report a retrospectively identified cohort of 23 neonates with electroencephalographically confirmed seizures who received levetiracetam. Levetiracetam was considered effective if administration was associated with a greater than 50% seizure reduction within 24 hours. Levetiracetam was initiated at a mean conceptional age of 41 weeks. The mean initial dose was 16 ± 6 mg/kg and the mean maximum dose was 45 ± 19 mg/kg/day. No respiratory or cardiovascular adverse effects were reported or detected. Levetiracetam was associated with a greater than 50% seizure reduction in 35% (8 of 23), including seizure termination in 7. Further study is warranted to determine optimal levetiracetam dosing in neonates and to compare efficacy with other antiseizure medications.
neonatal seizures; status epilepticus; levetiracetam; anticonvulsant
The purpose of this study was to assess deep hypothermic circulatory arrest (DHCA) as a modifier of neurodevelopmental (ND) outcomes in preschool children after cardiac surgery in infancy for repair of congenital heart defects (CHD).
This is a planned analysis of infants enrolled in a prospective study of apolipoprotein E polymorphisms and ND outcome after cardiac surgery. The effect of DHCA was assessed in patients with single or biventricular CHD without aortic arch obstruction. Neurodevelopmental assessment at 4 years of age included cognition, language, attention, impulsivity, executive function, social competence, and visual-motor and fine-motor skills. Patient and procedural variables were evaluated in univariate and multivariate models.
Neurodevelopmental testing was completed in 238 of 307 eligible patients (78%). Deep hypothermic circulatory arrest was used at the discretion of the surgeon at least once in 92 infants (38.6%) with a median cumulative duration of 36 minutes (range, 1 to 132 minutes). By univariate analysis, DHCA patients were more likely to have single-ventricle CHD (p = 0.013), lower socioeconomic status (p < 0.001), a higher incidence of preoperative ventilation (p < 0.001), and were younger and smaller at the first surgery (p < 0.001). By multivariate analysis, use of DHCA was not predictive of worse performance for any ND outcome.
In this cohort of children undergoing repair of CHD in infancy, patients who underwent DHCA had risk factors associated with worse ND outcomes. Despite these, use of DHCA for repair of single-ventricle and biventricular CHD without aortic arch obstruction was not predictive of worse performance for any ND domain tested at 4 years of age.
To determine whether a cardiac diagnosis is a predictor of neurodevelopmental outcomes after infant cardiac surgery.
Infants with ventricular septal defect (VSD), tetralogy of Fallot (TOF), transposition of the great arteries (TGA), and hypoplastic left heart syndrome (HLHS) in a study of apolipoprotein E (APOE) polymorphisms, and neurodevelopmental outcome underwent neurodevelopmental and genetic evaluation at 4 years of age. The domains tested included cognition, language, speech, memory, executive function, visual-motor, fine motor, and reading and math skills.
Testing was completed in 178 patients with normal genetic evaluations: VSD (n = 26), TOF (n = 44), TGA (n = 41), and HLHS (n = 67). No differences were found in gestational age, ethnicity, APOE genotype, socioeconomic status, or maternal education among groups. Patient age at the first surgery was significantly lower for patients with TGA and HLHS compared with those with TOF and VSD. The postoperative length of stay was significantly longer for HLHS than all other groups and for TGA compared with TOF and VSD. HLHS correlated significantly with the use of deep hypothermic circulatory arrest and multiple operations. The mean scores for each domain were within normal limits for all groups. Compared with the other patients, those with HLHS had significantly lower scores for cognition, fine motor skills, executive function, and math skills. No significant differences were found among the TGA, TOF or VSD patients for any domain. Significant impairments in at least 1 domain were identified in 8% (2/25) of patients with VSD, 20% (8/41) with TOF, 17% (7/41) with TGA, and 18% (12/65) with HLHS. After correction for the demographic, preoperative, and operative variables, no significant differences were found among the groups for any domain.
The mean scores for the neurodevelopmental outcomes domains tested were in the normal range for preschool children with no recognized genetic syndromes after surgery for VSD, TOF, TGA, and HLHS. In each diagnostic group, the number of children with impairments in at least 1 domain increased compared with the general population. Unadjusted neurodevelopmental outcomes for HLHS were lower for cognition, fine motor skills, executive function, and math skills compared with the other patients. After correction for the demographic, preoperative, and operative variables, no significant differences were found among the groups for any domain. The specific cardiac diagnosis determines a large portion of the variation in these covariates. Therefore, although HLHS did predict for poorer outcomes in some domains, it did not add predictive power to the other factors considered.
Electroencephalographic (EEG) features may provide objective data regarding prognosis in children resuscitated from cardiac arrest (CA), but therapeutic hypothermia (TH) may impact its predictive value. We aimed to determine whether specific EEG features were predictive of short-term outcome in children treated with TH after CA, both during hypothermia and after return to normothermia.
Thirty-five children managed with a standard clinical TH algorithm after CA were prospectively enrolled. EEG recordings were scored in a standardized manner and categorized. EEG category 1 consisted of continuous and reactive tracings. EEG category 2 consisted of continuous but unreactive tracings. EEG category 3 included those with any degree of discontinuity, burst suppression, or lack of cerebral activity. The primary outcome was unfavorable short-term outcome defined as Pediatric Cerebral Performance Category score of 4–6 (severe disability, vegetative, death) at hospital discharge. Univariate analyses of the association between EEG category and outcome was performed using logistic regression.
For tracings obtained during hypothermia, patients with EEGs in categories 2 or 3 were far more likely to have poor outcome than those in category 1 (OR 10.7, P = 0.023 and OR 35, P = 0.004, respectively). Similarly, for tracings obtained during normothermia, patients with EEGs in categories 2 or 3 were far more likely to have poor outcomes than those in category 1 (OR 27, P = 0.006 and OR 18, P = 0.02, respectively).
A simple EEG classification scheme has predictive value for short-term outcome in children undergoing TH after CA.
Therapeutic hypothermia; Outcome; Pediatric; Hypoxic ischemic encephalopathy; Heart arrest; Prognosis
Correct outcome prediction after cardiac arrest in children may improve clinical decision making and family counseling. Various investigators have used EEG to predict outcome with varying success, but one limiting issue is the potential lack of reproducibility of EEG interpretation. Therefore, we aimed to evaluate interobserver agreement using standardized terminology in the interpretation of EEG tracings obtained from critically ill children following cardiac arrest.
3 pediatric neurophysiologists scored 74 EEG samples using standardized categories, terminology, and interpretation rules. Interobserver agreement was evaluated using kappa and intra-class correlation coefficients.
Agreement was substantial for the categories of continuity, burst suppression, sleep architecture, and overall rating. Agreement was moderate for seizure occurrence and inter-ictal epileptiform discharge type. Agreement was fair for inter-ictal epileptiform discharge presence, beta activity, predominant frequency, and fastest frequency. Agreement was slight for maximum voltage and focal slowing presence.
The variability of inter-rater agreement suggests that some EEG features are superior to others for use in a predictive algorithm. Using only reproducible EEG features is needed to ensure the most accurate and consistent predictions. Since even seizure identification had only moderate agreement, studies of non-convulsive seizures in critically ill patients must be conducted and interpreted cautiously.
Electroencephalogram; Interobserver variability; Seizure; Pediatric; Hypoxic Ischemic Encephalopathy; Cardiac Arrest
To estimate the prevalence and identify the predictors of impaired growth following infant cardiac surgery.
Secondary analysis of a prospective study of the role of apolipoprotein E (APOE) gene polymorphisms on neurodevelopment in young children following infant cardiac surgery. Prevalence estimates for growth velocity were derived using anthropometric measures [weight (WT) and head circumference (HC)] obtained at birth and at 4-years of age. Genetic evaluation was also performed. Growth measure z-scores were calculated using World Health Organization Child Growth Standards. Growth velocity was evaluated using two different techniques: first by clustering the children into one of three growth velocity subgroups based on z-score: impaired growth (difference < − 0.5 SD), stable growth (difference of −0.5 SD to 0.5 SD), and growth improving (difference > 0.5 SD), and, second, using continuous difference scores. Statistical analyses were conducted using a combination of proportional odds models for the ordered categories and simple linear regression for the continuous outcomes.
Three hundred and nineteen full term subjects had complete anthropometric measures for WT and HC at birth and at 4-yrs. The cohort was 56% male. Genetic examinations were available for 97% (309/319) of the cohort (normal, 74% definite or suspected genetic abnormality, 26%). Frequency counts for WT categories were: impaired growth 37% stable growth 31% and improving growth 32%. Frequency counts for HC categories were: impaired growth 39% stable growth 28% and improving growth 33%. Presence of a definite or suspected genetic syndrome (p = 0.04) was found to be a predictor of impaired growth for WT, but not HC. When growth z-scores were used as continuous outcomes, the APOE ε2 allele was found to be predictive of lower z-scores for both WT (p = 0.02) and HC (p = 0.03).
Impaired growth for both WT and HC is common (both > 30%) in this cohort of children following infant cardiac surgery. Both the APOE ε2 allele and the presence of a definite or suspected genetic syndrome were associated with impaired WT growth velocity. The APOE ε2 allele was also associated with impaired growth velocity for HC. Persistent poor growth may have long-term implications for the health and development of children with CHD.
Heart Defects, Congenital; Genetic Predisposition to Disease; Apolipoproteins E; Growth Impairment
Hyperglycaemia has been associated with worse outcome following traumatic brain injury and cardiac surgery in adults. We have previously reported no relationship between early postoperative hyperglycaemia and worse neurodevelopmental outcome at 1 year following biventricular repair of congenital heart disease. It is not known if postoperative hyperglycaemia results in worse neurodevelopmental outcome after infant cardiac surgery for single-ventricle lesions.
Secondary analysis of postoperative glucose levels in infants <6 months of age undergoing Stage I palliation for various forms of single ventricle with arch obstruction. The patients were enrolled in a prospective study of genetic polymorphisms and neurodevelopmental outcomes assessed at 1 year of age with the Bayley Scales of Infant Development-II yielding two indices: mental developmental index (MDI) and psychomotor developmental index (PDI).
Stage I palliation was performed on 162 infants with 13 hospital and 15 late deaths (17.3% 1-year mortality). Neurodevelopmental evaluation was performed in 89 of 134 (66.4%) survivors. Glucose levels at admission to the cardiac intensive care unit and during the first 48 postoperative hours were available for 85 of 89 (96%) patients. Mean admission glucose value was 274 ± 91 mg dl−1; the maximum was 291 ± 90 mg dl−1, with 69 of 85 (81%) patients having at least one glucose value >200 mg dl−1. Only two patients had a value <50 mg dl−1. Mean MDI and PDI scores were 88 ± 16 and 71 ± 18, respectively. There were no statistically significant correlations between initial, mean, minimum or maximum glucose measurements and MDI or PDI scores. Only delayed sternal closure resulted in a statistically significant relationship between initial, minimum and maximum glucose values within the context of a multivariate analysis of variance model.
Hyperglycaemia following Stage I palliation in the neonatal period was not associated with lower MDI or PDI scores at 1 year of age.
Congenital heart disease; Hyperglycaemia; Patient outcomes; Postoperative care
The prevalence of perioperative stroke in infants undergoing operations for congenital heart disease has not been well described. The objectives of this study were to determine the prevalence of stroke as assessed by postoperative brain magnetic resonance imaging (MRI), characterize the neuroanatomic features of focal ischemic injury, and identify risk factors for its development.
Brain MRI was performed in 122 infants 3 to 14 days after cardiac operation with cardiopulmonary bypass, with or without deep hypothermic circulatory arrest. Preoperative, intraoperative, and postoperative data were collected. Risk factors were tested by logistic regression for univariate and multivariate associations with stroke.
Stroke was identified in 12 of 122 patients (10%). Strokes were preoperative in 6 patients and possibly intraoperative or postoperative in the other 6 patients, and were clinically silent except in 1 patient who had clinical seizures. Arterial-occlusive and watershed infarcts were identified with equal distribution in both hemispheres. Multivariate analysis identified lower birth weight, preoperative intubation, lower intraoperative hematocrit, and higher blood pressure at admission to the cardiac intensive care unit postoperatively as significant factors associated with stroke. Prematurity, younger age at operation, duration of cardiopulmonary bypass, and use of deep hypothermic circulatory arrest were not significantly associated with stroke.
The prevalence of stroke in infants undergoing operations for congenital heart disease was 10%, half of which occurred preoperatively. Most were clinically silent and undetected without neuroimaging. Mechanisms included thromboembolism and hypoperfusion, with patient-specific, procedure-specific, and postoperative contributions to increased risk.
The goal was to evaluate polymorphisms of the APOE gene as modifiers of neurobehavioral outcomes for preschool-aged children with congenital heart defects, after cardiac surgery.
A prospective observational study with neurodevelopmental evaluation between the fourth and fifth birthdays was performed. Attention and behavioral skills were assessed through parental report.
Parents of 380 children completed the neurobehavioral measures. Child Behavior Checklist scores for the pervasive developmental problem scale were in the at-risk or clinically significant range for 15% of the cohort, compared with 9% for the normative data (P < .00001). Attention problem scores were in the at-risk or clinically significant range for 12% of the cohort, compared with 7% for the normative data (P = .0002). The Attention-Deficit/Hyperactivity Disorder Rating Scale-IV, Preschool Version, was completed for 378 children; 30% scored in the clinically significant range for inattention and 22% for impulsivity. After adjustment for covariates, the APOE ε2 allele was significantly associated with higher scores (worse problems) for multiple Child Behavior Checklist indices, including somatic complaints (P = .009), pervasive developmental problems (P = .032), and internalizing problems (P=.009). In each case, the ε4 allele was associated with a better outcome. APOE ε2 carriers had impaired social skills, compared with ε4 carriers (P = .009).
For preschool-aged children with congenital heart defects requiring surgery, parental rating scales showed an increased prevalence of restricted behavior patterns, inattention, and impaired social interactions. The APOE ε2 allele was associated with increased behavior problems, impaired social interactions, and restricted behavior patterns.
congenital heart defects; genetic predisposition to disease; apolipoprotein E; behavioral symptoms; attention-deficit/hyperactivity disorder; impulsive behavior; autistic disorder
Many studies of neurodevelopmental outcomes after neonatal and infant cardiac surgery have focused on potentially modifiable risk factors for adverse outcomes, primarily intraoperative management strategies and the use of deep hypothermic circulatory arrest. There is increasing evidence that patient-specific factors are more important determinants of outcome.
We investigated predictors of neurodevelopmental outcomes at 1 year of age after neonatal and infant cardiac surgery in a subgroup of infants enrolled in a prospective study of apolipoprotein E (APOE) genotype and neurodevelopmental outcome. Children with a variety of 2-ventricle cardiac defects repaired with only 1 operation with cardiopulmonary bypass and no more than 1 episode of deep hypothermic circulatory arrest were included. Neurodevelopmental outcomes at 1 year of age included the Bayley Scales of Infant Development-II, which yield 2 indices, the Mental Developmental Index and the Psychomotor Developmental Index.
Two hundred forty-seven infants underwent surgical repair between October 1998 and April 2003 with 1 hospital death and 3 deaths before 1 year of age. Neurodevelopmental evaluation was performed in 188 (77%) of 243 survivors, including 56 patients with tetralogy of Fallot, 39 with transposition of the great arteries with intact ventricular septum, 34 with ventricular septal defects, and 59 with other defects. The median age at operation was 56 days (1–186 days), including 72 (38%) neonates. Confirmed or suspected genetic syndromes were present in 59 (31%) of 188 infants. Deep hypothermic circulatory arrest was used in 67 (35%) infants with a median duration of 34 minutes (1–80 minutes). For the entire cohort, the mean Mental Developmental Index was 90.6 ± 14.9 and the mean Psychomotor Developmental Index was 81.6 ± 17.2. For patients without genetic syndromes, the mean Mental Developmental Index was 93.7 ± 13.6 and the mean Psychomotor Developmental Index was 85.1 ± 14.6. For the entire cohort, predictors of lower scores for both the Mental Developmental Index and Psychomotor Developmental Index were presence of a confirmed or suspected genetic syndrome, lower birth weight, and presence of the APOE ε2 allele (all P < .04). Black race was associated with higher scores on the Psychomotor Developmental Index (P = .018). Lower nasopharyngeal temperature during cardiopulmonary bypass was associated with a lower score on the Psychomotor Developmental Index (P = .03) and was the only intraoperative factor that was a significant predictor of either the Mental or Psychomotor Developmental Index.
The strongest predictors of a worse neurodevelopmental outcome at 1 year of age were patient-specific factors including presence of a genetic syndrome, low birth weight, and presence of the APOE ε2 allele. Patient-specific factors eclipsed the use and duration of deep hypothermic circulatory arrest as predictors of worse neurodevelopmental outcomes.
The occurrence of a seizure after the arterial switch operation is associated with a worse long-term neurodevelopmental outcome. The significance of seizures after neonatal and infant repair of other congenital heart defects is not known.
A recent study at our institution demonstrated seizures documented by 48-hour electroencephalographic monitoring in 20 (11%) of 178 neonates and infants after surgery for complex congenital heart defects, including hypoplastic left heart syndrome or variants. The developmental outcomes of this cohort were evaluated at 1 year of age by using the Bayley Scales of Infant Development II, which yields 2 scores: the Mental Developmental Index and the Psychomotor Developmental Index.
Developmental evaluations were performed in 114 (70%) of 164 survivors, including 36 with hypoplastic left heart syndrome. Postoperative electroencephalographic seizures had occurred in 15 (13%) of 114 of the entire group and in 8 (22%) of 36 of those with hypoplastic left heart syndrome. For the entire cohort, the Mental Developmental Index was 92.3 ± 13.5, and the Psychomotor Developmental Index was 79.9 ± 18.8 for patients without seizures, compared with 90.3 ± 10.7 and 74.4 ± 19.3 for those with seizures (both P > .5). For the hypoplastic left heart syndrome subgroup, the Mental Developmental Index was 92.3 ± 14.9, and the Psychomotor Developmental Index was 74.8 ± 19.3 for patients with seizures, compared with 91.9 ± 12.4 and 73.9 ± 18.3 for those without seizures (both P > .5). A frontal onset of seizures was predictive of a lower score on the Psychomotor Developmental Index, but not on the Mental Developmental Index.
The occurrence of a seizure after cardiac operation is a marker of central nervous system injury. However, in this cohort of neonates and infants with complex congenital heart defects, the occurrence of a seizure was not predictive of a worse developmental outcome at 1 year of age as assessed by the Bayley Scales of Infant Development II.
Electroencephalographic seizures have been shown to occur in 5% to 20% of neonates and infants after biventricular repair of a variety of cardiac defects. Occurrence of a seizure is a predictor of adverse long-term neurodevelopmental sequelae. The contemporary incidence of postoperative seizures after repair of cardiac defects such as hypoplastic left heart syndrome and other forms of single ventricle is not known.
A prospective study of 178 patients less than 6 months of age undergoing cardiopulmonary bypass with or without deep hypothermic circulatory arrest (DHCA) was conducted at a single institution from September 2001 through March 2003 to identify postoperative seizures assessed by 48-hour continuous video electroencephalographic monitoring.
Cardiac defects included transposition of the great arteries with or without a ventricular septal defect (n = 12), ventricular septal defect with or without coarctation (n = 28), tetralogy of Fallot (n = 24), hypoplastic left heart syndrome or variant (n = 60), other functional single ventricle (n = 14), and other defects suitable for biventricular repair (n = 40). Median age at the time of the operation was 7 days (range, 1–188 days) and was 30 days or less in 110 (62%) patients. DHCA was used in 117 (66%) patients, with multiple episodes in 9 patients. Median total duration of DHCA was 40 minutes (range, 1–90 minutes). Electroencephalographic seizures were identified in 20 (11.2%) patients. Seizures occurred in 15 (14%) of 110 neonates and 5 (7%) of 68 older infants. Seizures occurred in 1 (4%) of 24 patients with tetralogy of Fallot, 1 (8%) of 12 with transposition of the great arteries, and 11 (18%) of 60 with hypoplastic left heart syndrome or variant. By stepwise logistic regression analysis, once increasing duration of total DHCA (P = .001) was considered, no other variable improved prediction of occurrence of a seizure. Patients with DHCA duration of more than 40 minutes had an increased incidence of seizures (14/58 [24.1%]) compared with those with a DHCA duration of 40 minutes or less (4/59 [6.8%], P = .04). The incidence of seizures for patients with a DHCA duration of 40 minutes or less was not significantly different from those in whom DHCA was not used (2/61 [3.3%], P = .38).
In the current era, continuous electroencephalographic monitoring demonstrates early postoperative seizures in 11.2% of a heterogeneous cohort of neonates and infants with complex congenital heart defects. Increasing duration of DHCA was identified as a predictor of seizures. However, the incidence of seizures in children with limited duration of DHCA was similar to that in infants undergoing continuous cardiopulmonary bypass alone.
Periventricular leukomalacia (PVL) is part of a spectrum of cerebral white matter injury which is associated with adverse neurodevelopmental outcome in preterm infants. While PVL is common in neonates with cardiac disease, both before and after surgery, it is less common in older infants with cardiac disease. Pre-, intra-, and postoperative risk factors for the occurrence of PVL are poorly understood. The main objective of the present work is to identify potential hemodynamic risk factors for PVL occurrence in neonates with complex heart disease using logistic regression analysis and decision tree algorithms.
The postoperative hemodynamic and arterial blood gas data (monitoring variables) collected in the cardiac intensive care unit of Children's Hospital of Philadelphia were used for predicting the occurrence of PVL. Three categories of datasets for 103 infants and neonates were used-—(1) original data without any preprocessing, (2) partial data keeping the admission, the maximum and the minimum values of the monitoring variables, and (3) extracted dataset of statistical features. The datasets were used as inputs for forward stepwise logistic regression to select the most significant variables as predictors. The selected features were then used as inputs to the decision tree induction algorithm for generating easily interpretable rules for prediction of PVL.
Three sets of data were analyzed in SPSS for identifying statistically significant predictors (p < 0.05) of PVL through stepwise logistic regression and their correlations. The classification success of the Case 3 dataset of extracted statistical features was best with sensitivity (SN), specificity (SP) and accuracy (AC) of 87, 88 and 87%, respectively. The identified features, when used with decision tree algorithms, gave SN, SP and AC of 90, 97 and 94% in training and 73, 58 and 65% in test. The identified variables in Case 3 dataset mainly included blood pressure, both systolic and diastolic, partial pressures pO2 and pCO2, and their statistical features like average, variance, skewness (a measure of asymmetry) and kurtosis (a measure of abrupt changes). Rules for prediction of PVL were generated automatically through the decision tree algorithms.
The proposed approach combines the advantages of statistical approach (regression analysis) and data mining techniques (decision tree) for generation of easily interpretable rules for PVL prediction. The present work extends an earlier research [Galli KK, Zimmerman RA, Jarvik GP, Wernovsky G, Kuijpers M, Clancy RR, et al. Periventricular leukomalacia is common after cardiac surgery. J Thorac Cardiovasc Surg 2004;127:692–704] in the form of expanding the feature set, identifying additional prognostic factors (namely pCO2) emphasizing the temporal variations in addition to upper or lower values, and generating decision rules. The Case 3 dataset was further investigated in Part II for feature selection through computational intelligence.
Congenital heart disease; Data mining; Decision tree algorithms; Logistic regression; Prognostics; Periventricular leukomalacia
The objective of Part II is to analyze the dataset of extracted hemodynamic features (Case 3 of Part I) through computational intelligence (CI) techniques for identification of potential prognostic factors for periventricular leukomalacia (PVL) occurrence in neonates with congenital heart disease.
The extracted features (Case 3 dataset of Part I) were used as inputs to CI based classifiers, namely, multi-layer perceptron (MLP) and probabilistic neural network (PNN) in combination with genetic algorithms (GA) for selection of the most suitable features predicting the occurrence of PVL. The selected features were next used as inputs to a decision tree (DT) algorithm for generating easily interpretable rules of PVL prediction.
Prediction performance for two CI based classifiers, MLP and PNN coupled with GA are presented for different number of selected features. The best prediction performances were achieved with 6 and 7 selected features. The prediction success was 100% in training and the best ranges of sensitivity (SN), specificity (SP) and accuracy (AC) in test were 60-73%, 74-84% and 71-74%, respectively. The identified features when used with the DTalgorithm gave best SN, SP and AC in the ranges of 87-90% in training and 80-87%, 74-79% and 79-82% in test. Among the variables selected in CI, systolic and diastolic blood pressures, and pCO2 figured prominently similar to Part I. Decision tree based rules for prediction of PVL occurrence were obtained using the CI selected features.
The proposed approach combines the generalization capability of CI based feature selection approach and generation of easily interpretable classification rules of the decision tree. The combination of CI techniques with DT gave substantially better test prediction performance than using CI and DT separately.
Congenital heart disease; Computational intelligence; Data mining; Decision tree; Genetic algorithms; Neural networks; Periventricular leukomalacia
Small head circumferences and white matter injury in the form of periventricular leukomalacia have been observed in populations of infants with severe forms of congenital heart defects. This study tests the hypothesis that congenital heart defects delay in utero structural brain development.
Full-term infants with hypoplastic left heart syndrome or transposition of the great arteries were prospectively evaluated with preoperative brain magnetic resonance imaging. Patients with independent risk factors for abnormal brain development (shock, end-organ injury, or intrauterine growth retardation) were excluded. Outcome measures included head circumferences and the total maturation score on magnetic resonance imaging. Total maturation score is a previously validated semiquantitative anatomic scoring system used to assess whole brain maturity. The total maturation score evaluates 4 parameters of maturity: (1) myelination, (2) cortical infolding, (3) involution of glial cell migration bands, and (4) presence of germinal matrix tissue.
The study cohort included 29 neonates with hypoplastic left heart syndrome and 13 neonates with transposition of the great arteries at a mean gestational age of 38.9 ± 1.1 weeks. Mean head circumference was 1 standard deviation below normal. The mean total maturation score for the cohort was 10.15 ± 0.94, significantly lower than reported normative data in infants without congenital heart defects, corresponding to a delay of 1 month in structural brain development.
Before surgery, term infants with hypoplastic left heart syndrome and transposition of the great arteries have brains that are smaller and structurally less mature than expected. This delay in brain development may foster susceptibility to periventricular leukomalacia in the preoperative, intraoperative, and postoperative periods.