Neurologic events in neonates treated surgically for congenital heart disease

Transcription

1 ORIGINAL ARTICLE Neurologic events in neonates treated surgically for congenital heart disease VY Chock 1, VM Reddy 2, D Bernstein 1 and A Madan 1 (2006) 26, r 2006 Nature Publishing Group All rights reserved /06 $ Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA and 2 Cardiovascular Surgery, Stanford University School of Medicine, Stanford, CA, USA Objective: The incidence of acute neurologic events prior to discharge in neonates with congenital heart disease (CHD) was determined and peri-operative characteristics predictive of a neurologic event were identified. Study design: A retrospective chart review over 1 year was conducted of infants <1 month of age with a diagnosis of CHD. Outcomes were measured by the occurrence of an acute neurologic event defined as electroencephalogram (EEG)-proven seizure activity, significant hypertonia or hypotonia, or choreoathetosis prior to hospital discharge. Stepwise logistic regression identified variables most likely to be associated with an acute neurologic event. Results: Surgical intervention occurred in 95 infants who were admitted with a diagnosis of CHD. The survival rate was 92%. Of the survivors, 16 (17%) had an acute neurologic event, with 19% of events occurring preoperatively. Factors associated with neurologic events included an elevated nucleated red blood cell (NRBC) count, an abnormal preoperative brain imaging study, and a 5-min Apgar score <7 (P<0.05). Conclusions: Neonates with CHD have a significant risk of neurologic events. Preoperative brain imaging, the 5-min Apgar score, and initial serum NRBC counts may identify infants at highest risk for central nervous system injury. (2006) 26, doi: /sj.jp ; published online 23 February 2006 Keywords: cardiac surgery; neurologic injury; risk factors Introduction Congenital heart disease (CHD) requiring surgical intervention in infancy accounts for cases per 1000 live births. 1,2 Survivors are at higher risk for significant neurological deficits or developmental disabilities including impaired visual-motor skills Correspondence: Dr A Madan, Department of Pediatrics, Division of Neonatal and Developmental Medicine, 750 Welch Road, Suite 315, Palo Alto, CA 94304, USA. ashima@stanford.edu Received 7 September 2005; revised 9 January 2006; accepted 17 January 2006; published online 23 February 2006 and behavioral abnormalities at school age. 1,3,4 Various studies have reported that infants who undergo cardiac surgery are at risk for developing postoperative seizures, ischemic lesions on magnetic resonance imaging (MRI), and abnormal neurological examinations 1 year after surgery. 4 7 The timing and mechanism of neurological injury in these infants remain unclear. Injury may occur preoperatively, intraoperatively or postoperatively. 8,9 Preoperative injury may be caused by hemodynamic instability with profound circulatory failure or coexisting brain malformations. 10,11 Intraoperative global brain injury can occur in infants who undergo either deep hypothermia 12 or more than min of circulatory arrest. 13,14 In addition, exposure of blood to the cardiopulmonary bypass circuit may further trigger an inflammatory response and disrupt cerebral autoregulation. 15 Postoperative neurological injury can occur from hemodynamic instability, embolic events, severe hypoxemia or intracranial hemorrhage. Several studies have attempted to identify risk factors and radiologic or serum markers that can lead to early detection of neurological injury. Preoperative cranial ultrasound abnormalities were identified in 59% 11 and MRI abnormalities were detected in 24% of infants with a diagnosis of CHD. 7 Other studies have looked at specific subpopulations of infants with CHD, for example those with d-transposition of the great arteries (d-tga) or single ventricle physiology and identified factors such as lack of a prenatal diagnosis, length of deep hypothermic circulatory arrest, age at surgery, microcephaly, presence of an acyanotic heart lesion, genetic conditions, elevated serum lactate, and metabolic acidosis, to be associated with an increased risk of neurological injury in the immediate postoperative period and persistent developmental deficits several years after surgery. 4,6,16 18 This study was conducted at a major referral center that does not routinely use circulatory arrest during surgical repair. Owing to the recent expansion of our cardiovascular program, we were particularly interested in neurological outcomes since the initiation of a low flow cardiopulmonary bypass technique. Therefore, we determined the incidence of neurological events and the perioperative characteristics predictive of a neurological event prior

2 Neurologic injury in congenital heart disease 238 to discharge in all neonates requiring cardiac surgery over a period of 1 year. Methods A 1-year retrospective chart review was performed on all 124 infants less than 1 month of age with a diagnosis of CHD who were admitted to the neonatal or cardiac intensive care units for the period 3/1/02 to 2/28/03. The Institutional Review Board approved this medical record review. Analysis of demographic variables included gender, inborn status, prenatal diagnosis, gestational age, birth weight, age at admission, 5-min Apgar score, karyotype, and the presence of non-cardiac congenital anomalies. The type of cardiac lesion was classified into groups based on relative pulmonary and systemic blood flow. 19 Maternal data included Group B Streptococcus culture status and presence of prolonged rupture of membranes. Laboratory data included C-reactive protein, liver enzymes, the lowest arterial ph, highest base deficit, and highest lactate level in the first 24 h after admission. Absolute nucleated red blood cell (NRBC) count was recorded if obtained within the first 2 days of life. Pre- or postoperative brain imaging results by cranial ultrasound, MRI, or computed tomography scan were also analyzed. Preoperative cranial ultrasounds were routinely obtained in infants with CHD, while postoperative imaging was carried out for known brain abnormalities or for clinical indications. Examination of intraoperative variables included days until surgical repair, cardiopulmonary bypass time, aortic crossclamp time, and the use of deep hypothermia (cooling to <181C). Infants who did not undergo surgical repair due to futility of surgical correction (N ¼ 3) or who had primary repair by catheterization (N ¼ 26) (e.g. balloon valvuloplasty for critical pulmonary stenosis) were excluded. After these exclusion criteria were applied, a total of 95 infants remained in the study group. The primary outcome measure was the presence of an acute neurological event prior to hospital discharge. This was defined as electroencephalogram (EEG)-proven seizure activity or significant hypertonia, hypotonia or choreoathetosis as documented by a neurologist or developmental specialist. Tone was assessed by neurologic examination near discharge when infants were hemodynamically stable, off the ventilator, and non-sedated. Electroencephalograms were obtained if clinically indicated by the presence of seizure activity or unexplained vital sign instability as witnessed by medical personnel. Hypotonia in conjunction with Trisomy 21 was not considered a neurological event. Mortality was an additional outcome, defined as death within 30 days of surgical repair. Statistical analysis included coding of variables by a method similar to the Score for Neonatal Acute Physiology (SNAP) model 20 in which missing data were treated as normal for reference range for categorical variables and as the population median value for continuous variables. Several continuous variables were analyzed as categorical variables at predetermined normal values. Specifically, we looked for the presence of CRP>1.0 mg/dl, AST>80 U/l or ALT>50 U/l, 5-min Apgar <7, lactate >5 mmol/l, serum ph<7.0, base deficit < 10 mmol/l, and absolute NRBC count >2000. Univariate and multivariate analysis using stepwise forward logistic regression was done with Statview. A P-value<0.05 was considered statistically significant. Results Clinical characteristics There were almost equal numbers of male (51) and female (44) infants. The median gestational age was 39 weeks (range 26 to 42 weeks) and the median birth weight was 3098 g (range 1026 to 4621 g). Of the inborn infants, 77% had a prenatal diagnosis of cardiac disease compared to 13% of the outborn infants. Further study group characteristics and a univariate analysis of the perioperative factors associated with the occurrence of neurologic events are outlined in Appendix A of Supplementary Information. Appendix B of Supplementary Information summarizes the various diagnostic categories of CHD. Nearly half of the study group (46%) had a cardiac lesion with decreased systemic blood flow such as hypoplastic left ventricle, coarctation of the aorta, or an interrupted aortic arch. Chromosome results were available in 85 infants (89%). Chromosomal abnormalities (11%) included neonates with Trisomy 21 (n ¼ 2), 22q11 deletion (n ¼ 5), and one infant each with 45XO, 6p24 or 5p 21q chromosomes. Congenital anomalies (9%) included infants with VATER association (vertebral defects, anal atresia, tracheoesophageal fistula, radial and renal dysplasia), CHARGE sequence (coloboma, heart disease, atresia choanae, retarded growth, genital anomalies, ear anomalies), cleft palate, imperforate anus, heterotaxy, polydactyly and methotrexate toxicity. Two of these infants with anomalies also had abnormal chromosomes. Preoperative brain imaging by cranial ultrasound was obtained in 84% (80) of patients. No imaging was obtained in a small number of patients (15) due to inaccessibility of an ultrasound technician prior to surgery. However, there were no differences in preoperative variables, intraoperative variables, or adverse neurological outcomes in this group compared to those infants that did receive a preoperative cranial ultrasound. Postoperative imaging by follow-up cranial ultrasound, MRI, or computed tomography scan was obtained in 47% (45) of patients, but only if a preoperative study was abnormal or if clinically indicated, for example due to persistent desaturations, a prolonged code event, seizure activity, or tone abnormalities. An EEG was done in 27% (26) patients, similarly for clinical indications. Intraoperative characteristics included a mean bypass time of 148±52 min and aortic cross-clamp time of 65±40 min. Deep hypothermia was utilized in 36% of patients. No patient underwent circulatory arrest.

3 Neurologic injury in congenital heart disease 239 Radiologic data Abnormal results on preoperative imaging were found in 15% (12/ 80) and seven out of 12 of these infants died (two) or had acute neurological events (five) prior to discharge. Postoperative imaging was abnormal in 42% (19/45). Imaging abnormalities included ventricular hemorrhage, hydrocephalus, structural abnormalities, and white matter injury seen on MRI. A total of five infants had congenital brain malformations including lissencephaly, agenesis or dysgenesis of the corpus callosum, Dandy Walker malformation or holoprosencephaly. Those infants who died were more likely to have congenital brain malformations whereas those with acute neurological events had hemorrhagic abnormalities or periventricular leukomalacia. Of the 12 patients with preoperative imaging abnormalities, six persisted or evolved postoperatively (mostly hemorrhage and ventriculomegaly), three resolved (mild hemorrhage or edema), and three did not receive follow-up imaging prior to discharge (congenital anomalies). New imaging abnormalities were found postoperatively in 13 patients, including infarct, hemorrhage, atrophy, ventriculomegaly, periventricular leukomalacia and venous sinus thrombosis. Outcomes Ninety-five infants in the study group underwent surgical intervention for CHD. The survival rate was 92%, and of the survivors, 16 (17%) had acute neurological events prior to discharge. Characterization of these events is provided in Table 1. Tone abnormalities were seen in 56% and seizures occurred in 38% of affected infants. In addition, neurological events occurred preoperatively in 19% of infants, within 1 week following surgery in 25%, and >1-week postoperatively in 56%. Neurological events occurring within 1 week of surgery were all seizures, while those occurring after 1 week of surgery were mostly tone abnormalities. Abnormal preoperative imaging studies were seen in 31% (5/16) and abnormal postoperative imaging in 69% (11/16) of infants with acute neurological events in comparison to 9% (7/79) and 10% (8/79) in those without acute neurological events. In addition, 81% (13) of affected infants underwent cardiopulmonary bypass during their surgery. Variables associated with an acute neurological event Using logistic regression, a multivariate analysis found three variables to be associated with the occurrence of neurological events prior to discharge (Table 2). These variables were an absolute NRBC count >2000 on admission (OR 7.05, 95% CI 1.42 to 35.0, P ¼ 0.02), an abnormal preoperative brain imaging study (OR 4.75, 95% CI 1.17 to 19.3, P ¼ 0.03), and a 5-min Apgar score <7 (OR 4.78, 95% CI 1.06 to 21.6, P ¼ 0.04). Absolute NRBC count was available in 72% of infants (68/95) and as a continuous variable was also significantly associated with the occurrence of neurological events (P ¼ 0.02). In a univariate analysis, the presence of congenital anomalies alone had a significant Table 1 Acute neurologic events Category N (%) (total N ¼ 16) Type of neurologic event Seizures 6 (38) Tone abnormalities 9 (56) Choreoathetosis 1 (6) Timing of neurologic event Preoperative 3 (19) Postoperative <1 week 4 (25) Postoperative >1 week 9 (56) Event with abnormal imaging Preoperative 5 (31) Postoperative 11 (69) Table 2 Multivariate analysis of perioperative factors associated with the occurrence of neurologic events Risk factor Odds ratio 95% Confidence interval P-value Absolute NRBC s > ( ) 0.02 Abnormal preoperative imaging 4.75 ( ) min Apgar < ( ) 0.04 association with neurological events (P ¼ 0.05), but was not significant when combined with other variables. All variables with their corresponding P-values are listed in Appendix A of Supplementary Information. Discussion For infants with CHD, improved surgical techniques and better perioperative management are likely to increase the number of survivors, as well as the number of infants with neurological deficits. Identification of the population at greatest risk for neurological morbidity will lead to better anticipatory guidance, close developmental follow-up, early intervention with rehabilitation services, and potential institution of neuroprotective strategies. In this study, we sought to characterize the at risk population. Our population was at higher risk for neurologic morbidity due to the inclusion of preterm infants and a large percentage of infants with decreased systemic blood flow from their cardiac lesion. However, in our analysis, prematurity and type of cardiac lesion did not appear to be variables associated with poor neurological outcome. In addition, none of our surgical repairs involved circulatory arrest, a technique that can cause cerebral hypoperfusion and a known risk factor for poor neurological outcomes. 16

4 240 Neurologic injury in congenital heart disease The 17% risk of neurological events prior to discharge in our high-risk group of infants is comparable to that of other studies. In a previous study of infants with hypoplastic left heart syndrome, a similar incidence of neurological injury (22%) was found. 17 In a separate study, infants with d-tga also had an increased incidence of neurological events. 16 Other studies that excluded survivors of hypoplastic left heart syndrome found clinical seizures or coma in 18.9% of infants postoperatively 6 and tone abnormalities, motor asymmetry, jitteriness or cranial nerve palsies in >50%. 9 These populations were predominantly infants undergoing palliation with deep hypothermic circulatory arrest. While circulatory arrest and its associated decrease in cerebral blood flow may contribute to worse neurological outcomes, our study suggests that other factors must be involved in generating neurological injury. A 17% risk for a neurological event in a neonate with CHD may be an underestimate. The neurological status of a critically ill patient given heavy sedation is inherently difficult to evaluate. We only accounted for EEG-proven seizures, but a significant number of ictal events may have been missed by EEG. Indeed, clinical descriptions of desaturations, posturing, and tonic clonic events were not always captured by EEG or had resolved prior to examination by a neurologist. It is unclear if the identified neurological abnormalities resolved after hospital discharge or if they had any permanent effect on neurodevelopmental outcome. From the Boston Circulatory Arrest study of infants with d-tga, evidence of seizures within 48 h of surgery was one of the strongest predictors of neurological abnormalities at 1 year of life and lower mean IQ scores at 4 years of age. 14,21 Limperopoulos also found that perioperative neurodevelopmental abnormalities were a marker of subsequent neurological, motor, and global deficits. 4 These studies indicate that in our cohort of patients, early neurological events are likely indicative of poor long-term outcome, although their significance would be enhanced with neurodevelopmental follow-up. In this study, a low 5-min Apgar, elevated NRBC count, and abnormalities on preoperative brain imaging were identified as variables associated with adverse neurological events. In previous studies, a low 5-min Apgar score was shown to be an independent risk factor for increased mortality 17 and neurological abnormalities at 1yearofage. 21 Although these Apgar scores are subject to bias, especially among infants given a prenatal diagnosis of CHD, low Apgar scores should identify those infants who are clinically unstable immediately after delivery and may require closer monitoring. Elevated NRBC counts could also be a useful marker for subsequent brain injury in this specific population of infants with CHD. Elevated NRBC counts are a marker of fetal hypoxia and may predict poor outcome after hypoxic ischemic encephalopathy, 22,23 although other studies have not found a correlation with brain injury. 24 There have not been any specific studies on the predictive value of NRBC counts in patients with CHD. Most studies evaluate NRBC counts in patients after perinatal depression and typically exclude infants with congenital cardiac malformations from their analyses. It is likely that infants with CHD also undergo chronic in utero hypoxia with a compensatory increase in erythropoiesis as estimated by the NRBC count after birth. Infants with left-sided obstructive lesions resulting in decreased brain oxygenation would be most at risk, although our study did not find type of cardiac lesion to be a significant predictor of injury or NRBC count. Our study may even underestimate the number of infants with elevated NRBC counts, as those infants transferred into our institution may have normalized their NRBC counts before accurate measurements could be made. To the best of our knowledge, no other study has been able to correlate abnormal brain imaging results with clinical neurological outcome. 4,21 The incidence of preoperative abnormalities has varied from 24 to 40% by MRI 7,25 to 59% by cranial ultrasound. 11 These abnormalities have included periventricular leukomalacia, infarct, parenchymal hemorrhage, and diffuse white matter injury. Interestingly, focal lesions identified by MRI were not correlated with clinical examination 25 and resolution of lesions was common 4 to 6 months after surgery. 7 Therefore, early radiologic findings are of unclear significance. Our analysis revealed a small number of patients with abnormal brain imaging preoperatively (15%). However, infants with an abnormal preoperative imaging study were much more likely to have a neurological event during their hospital stay. Later neurodevelopmental outcome data will be useful to confirm the relevance of early brain imaging studies. Our study found an association with preoperative variables and occurrence of neurological events, but not with intraoperative variables. The majority of infants did undergo cardiopulmonary bypass, but variables associated with this process were not significant. In addition, >50% of acute neurological events were detected >1-week postoperatively, a time period later than expected for embolic or hypoperfusion injuries related to cardiopulmonary bypass. Furthermore, circulatory arrest, a known intraoperative variable associated with adverse neurological sequelae, 3,4,16 was not utilized in our study. Recent improvements in operative technique and cardiopulmonary bypass may have reduced the contribution of intraoperative variables to neurological outcome. In contrast to our findings, other studies have implicated additional preoperative variables as predictors of neurological sequelae. The presence of a genetic syndrome or abnormal chromosomes, 6 metabolic acidosis or elevated serum lactate levels, 18 and specific cardiac lesions including acyanotic heart disease, aortic arch obstruction, or ventricular septal defect with d-tga 3,4,6 have been associated with neurological events. Although we chose to study a heterogeneous population of infants with all types of cardiac disease, our study did not implicate a specific type of heart lesion as more susceptible to neurological injury. Reclassification by individual lesion rather than by relative pulmonary and systemic blood flow may have yielded significant

5 Neurologic injury in congenital heart disease 241 results; however, we did not have sufficient numbers of infants to justify this approach. While the presence of congenital anomalies was significant as an independent variable in our study, when combined with other variables, it lost significance. Specifically, nearly half of infants (four) with congenital anomalies also had abnormal preoperative imaging. In addition, analysis of other variables including additional indicators of fetal distress, microcephaly, anemia, measures of cerebral perfusion and degree of hypocapnia may have yielded additional data. Although conclusions from this study are limited due to its retrospective design with a lack of laboratory data and consistent imaging available for all patients, it provides us with important data for the design of future prospective studies. Given that not all subjects had an EEG, the timing at which a neurological event was documented pre- or postoperatively was difficult to define. Neurologic events that may have occurred after hospital discharge were also not assessed. In addition, cardiorespiratory compromise and illness severity may have contributed to decreased neurological integrity. The prevalence of neuroimaging abnormalities may also be underestimated as infants with acquired injury or subtle congenital malformations may not demonstrate clinical manifestations and may not reveal injury on a cranial ultrasound alone. Despite these limitations, very few studies have specifically analyzed a population that did not undergo circulatory arrest during surgical repair. The persistence of neurological injury after elimination of this risk factor is notable and worthy of further study with a prospective trial design. Conclusion This study adds to the growing body of literature documenting the risk of acute neurological events both pre- and postoperatively in neonates with CHD. Risk factors for neurological events include an elevated NRBC count, low 5-min Apgar score, and abnormal preoperative brain imaging studies. No intraoperative variables were found to be significant in this analysis. The use of preoperative brain imaging by cranial ultrasound or MRI in all neonates with CHD would be useful to identify infants at high risk of neurological injury. In addition, it is important to have a high index of suspicion for neurological abnormalities and a plan for thorough neurodevelopmental follow-up in any cardiac patient with significant risk factors. The design of future prospective studies and implementation of neuroprotective strategies may especially benefit those infants with low Apgar scores, high NRBC values and abnormal brain imaging. Acknowledgments We thank William E Benitz, MD for his assistance with the statistical analysis and Stephen J Roth, MD for his critical review of the manuscript. Dr Chock received support from the Neonatal and Developmental Biology Training Grant at Stanford University. References 1 Ferry PC. Neurologic sequelae of cardiac surgery in children. AJDC 1987; 41: Ferry PC. Neurologic sequelae of open-heart surgery in children. AJDC 1990; 144: Bellinger DC, Wypij D, duplessis A, Rappaport LA, Jonas RA, Wernovsky G et al. Neurodevelopmental status at eight years in children with dextrotransposition of the great arteries: The Boston Circulatory Arrest Trial. J Thorac Cardiovasc Surg 2003; 126: Limperopoulos C, Majnemer A, Shevell MI, Rohlicek C, Rosenblatt B, Tchervenkov C et al. Predictors of developmental disabilities after open heart surgery in young children with congenital heart defects. J Pediatr 2002; 141: Miller G, Eggli KD, Contant C, Baylen BG, Myers JL. Postoperative neurologic complications after open heart surgery in young infants. Arch Pediatr Adolesc Med 1995; 149: Clancy RR, McGaurn SA, Wernovsky G, Gaynor JW, Spray TL, Norwood WI et al. Risk of seizures in survivors of newborn heart surgery using deep hypothermic circulatory arrest. Pediatrics 2003; 111: Mahle WT, Tavani F, Zimmerman R, Nicolson S, Galli KK, Gaynor JW et al. An MRI study of neurological injury before and after congenital heart surgery. Circulation 2002; 106: I109 I Limperopoulos C, Majnemer A, Shevell MI, Rosenblatt B, Rohlicek C, Tchervenkov C et al. Neurologic status of newborns with congenital heart defects before open heart surgery. Pediatrics 1999; 103: Limperopoulos C, Majnemer A, Shwevell MI, Rosenblatt B, Rohlicek C, Tchervenkov C et al. Neurodevelopmental status of newborns and infants with congenital heart defects before and after open heart surgery. J Pediatr 2000; 137: Glauser TA, Rorke LB, Weinberg PM, Clancy RR. Congenital brain anomalies associated with the hypoplastic left heart syndrome. Pediatrics 1990; 85: Van Houten JP, Rothman A, Bejar R. High incidence of cranial ultrasound abnormalities in full-term infants with congenital heart disease. Am J Perinatol 1996; 13: Dexter F, Hindman B. Theoretical analysis of cerebral venous blood hemoglobin oxygen saturation as an index of cerebral oxygenation during hypothermic cardiopulmonary bypass. Anesthesiology 1995; 83: Wypij D, Newburger JW, Rappaport LA, duplessis AJ, Jonas RA, Wernovsky G et al. The effect of duration of deep hypothermic circulatory arrest in infant heart surgery on late neurodevelopment: The Boston Circulatory Arrest Trial. J Thorac Cardiovasc Surg 2003; 126: Bellinger DC, Wypij D, Kuban KC, Rappaport LA, Hickey PR, Wernovsky G et al. Developmental and neurological status of children at 4 years of age after heart surgery with hypothermic circulatory arrest or low-flow cardiopulmonary bypass. Circulation 1999; 100: Steinberg JB, Kapelanski DP, Olson JD, Weiler JM. Cytokine and complement levels in patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1993; 106: Newburger JW, Jonas RA, Wernovsky G, Wypij D, Hickey PR, Kuban KC et al. A comparison of the perioperative neurologic effects of hypothermic circulatory arrest versus low-flow cardiopulmonary bypass in infant heart surgery. N Engl J Med 1993; 329:

6 242 Neurologic injury in congenital heart disease 17 Mahle WT, Clancy RR, McGaurn SP, Goin JE, Clark BJ. Impact of prenatal diagnosis on survival and early neurologic morbidity in neonates with the hypoplastic left heart syndrome. Pediatrics 2001; 107: Trittenwein G, Nardi A, Pansi H, Golej J, Burda G, Hermon M et al. Early postoperative prediction of cerebral damage after pediatric cardiac surgery. Annal Thorac Surg 2003; 76: Hamrick SE, Gremmels DB, Keet CA, Leonard CH, Connell JK, Hawgood S et al. Neurodevelopmental outcome of infants supported with extracorporeal membrane oxygenation after cardiac surgery. Pediatrics 2003; 111: e671 e Richardson DK, Gray JE, McCormick MC, Workman K, Goldmann DA. Score for Neonatal Acute Physiology: a physiologic severity index for neonatal intensive care. Pediatrics 1993; 91: Bellinger DC, Jonas R, Rappaport L, Wypij D, Wernovsky G, Kuban KC et al. Developmental and neurologic status of children after heart surgery with hypothermic circulatory arrest or low-flow cardiopulmonary bypass. N Engl J Med 1995; 332: Korst LM, Phelan JP, Ahn MO, Martin GI. Nucleated red blood cells: an update on the marker for fetal asphyxia. Am J Obstet Gynecol 1996; 175: Buonocore G, Perrone S, Gioia D, Gatti MG, Massafra C, Agosta R et al. Nucleated red blood cell count at birth as an index of perinatal damage. Am J Obstet Gynecol 1999; 181: Hamrick SE, Miller SP, Newton NR, Parer JT, Ferriero DM, Barkovich AJ et al. Nucleated red blood cell counts: not associated with brain injury or outcome. Pediatr Neurol 2003; 29: Miller SP, McQuillen PS, Vigneron DB, Glidden DV, Barkovich AJ, Ferriero DM et al. Preoperative brain injury in newborns with transposition of the great arteries. Ann Thorac Surg 2004; 77: Supplementary Information accompanies the paper on the website (