Journal of Perinatology (2011) 31, 392–396 r 2011 Nature America, Inc. All rights reserved. 0743-8346/11 www.nature.com/jp ORIGINAL ARTICLE Acute neurological findings in a national cohort of neonates with severe neonatal hyperbilirubinemia M Sgro1,2, D Campbell2, T Barozzino2 and V Shah3 1Keenan Research Centre of the Li Ka Shing Knowledge Institute, University of Toronto, Toronto, ON, Canada; 2Department of Pediatrics, St Michael’s Hospital, University of Toronto, Toronto, ON, Canada and 3Department of Pediatrics, Mount Sinai Hospital, University of Toronto, Toronto, ON, Canada Introduction Objective: To describe and compare infants with severe Neonatal hyperbilirubinemia remains the most common cause hyperbilirubinemia, who presented with and without abnormal neurological of readmission in the neonatal period, despite attempts to identify findings and to identify associated risk factors. at-risk newborns for severe hyperbilirubinemia before initial Study Design: Data on infants with severe hyperbilirubinemia discharge from the hospital.1–10 Long-term sequelae of severe (>425 mmol lÀ1 and/or received exchange transfusion) were collected hyperbilirubinemia, including bilirubin encephalopathy and prospectively through the Canadian Paediatric Surveillance Program kernicterus, were thought to be rare since the advent of maternal (CPSP) from 2002 to 2004. Cases were categorized into two groups on the rhesus immunoglobulin prophylaxis, phototherapy and exchange basis of information provided by the reporting physician: neurologically transfusions.11–13 However, cases of bilirubin encephalopathy and normal or abnormal. Demographic characteristics were compared and kernicterus continue to be reported in healthy near-term and term univariate logistic regression was performed to identify factors associated infants in North America and Europe.1,11–20 with acute neurological abnormalities in infants. We recently reported on 258 infants with severe Result: Of the initial cohort of 258 infants, 32 (12.4%) were identified to hyperbilirubinemia in Canada from 2002 to 2004 through the 10 have neurological abnormalities. Infants in the highest peak bilirubin CPSP. The most common identifiable etiology of severe level group (>550 mmol lÀ1) had the greatest risk of acute neurological hyperbilirubinemia in our cohort was ABO incompatibility followed abnormalities. The mid range (451 to 550 mmol lÀ1) and lowest level by glucose-6-phosphate dehydrogenase deficiency, with many (p450 mmol lÀ1) groups were less likely to have abnormalities (odds infants presenting with abnormal neurological symptoms and signs ratio (OR) ¼ 0.174; P ¼ 0.0013 and 0.402; P ¼ 0.0613, respectively). on admission. Exchange transfusion and presentation within the first 2 days of age were Acute bilirubin encephalopathy occurs when bilirubin levels 9,12 positively associated with abnormal neurological findings in infants become neurotoxic in infants. Three phases of acute bilirubin (OR ¼ 3.332, P ¼ 0.003 and OR ¼ 2.572, P<0.0001, respectively). encephalopathy have been described. The early acute phase is characterized by lethargy, hypotonia and poor suck; intermediate Conclusion: In this national cohort of infants with severe phase by hypertonia (including opisthotonus and retrocollis), hyperbilirubinemia, a significant percentage of infants developed acute high-pitched cry, fever, moderate stupor or irritability, whereas in bilirubin encephalopathy. Long-term neurodevelopmental follow-up is the advanced phase infants present with seizures, coma, apnea, necessary to determine the incidence of permanent neurological sequelae. pronounced retrocollis and opisthotonus, with death occurring in Journal of Perinatology (2011) 31, 392–396; doi:10.1038/jp.2010.137; severe cases.9,12 Although the acute neurological abnormalities of published online 9 December 2010 hyperbilirubinemic infants can be transient, the intermediate and Keywords: severe hyperbilirubinemia; bilirubin encephalopathy; advanced phases have been strongly linked to long-term infant-newborn neurological sequelae.21 The inability of physicians to recognize bilirubin encephalopathy early may lead to delays in treatment, thereby increasing the risk of long-term sequelae.22 The objective of this study was to describe and compare Correspondence: Dr D Campbell, Department of Pediatrics, St Michael’s Hospital, University the clinical course of infants with and without abnormal of Toronto, 30 Bond Street, Rm 15-014 Cardinal Carter, Toronto, ON M5B 1W8, Canada. neurological findings on admission to identify and better refine E-mail: [email protected] or [email protected] risk factors associated with the occurrence of acute neurological Received 10 June 2010; revised 11 August 2010; accepted 16 August 2010; published online 9 December 2010 findings. Neonatal hyperbilirubinemia and neurological findings M Sgro et al 393 Methods with acute bilirubin encephalopathy.12 Six of these infants Subjects were identified from a previously reported national had intermediate or advanced bilirubin encephalopathy database on infants with severe hyperbilirubinemia with seizures and/or opisthotonus during the initial (>425 mmol lÀ1 and/or received exchange transfusion) through presentation. the CPSP for the period July 2002 to June 2004.10 The methodology The baseline characteristics of the two cohorts (neonates with that was utilized to obtain data has been previously described.10 and without abnormal neurological findings) were similar The protocol and questionnaire were reviewed and approved by the (Table 1). The clinical presentation of the 32 infants with acute CPSP Steering Committee panel of experts, and by the Research neurological abnormalities is presented in Table 2. The peak Ethics Boards of two hospitals (St Michael’s Hospital and Mount bilirubin was higher in those who were neurologically abnormal À1 Sinai Hospital, Toronto, ON). A subsequent protocol was reviewed ((508.2 (104.5) vs 466.8 (71.8) mmol l )). Infants that presented and approved by the Research Ethics Boards of both institutions at <48 h of age had significantly lower serum bilirubin levels À1 for this secondary analysis. (median peak bilirubin ¼ 444 mmol l s.d. 88.6) compared Cases in which the reporting physician identified neurological with infants that presented after 48 h (median peak À1 abnormalities in the neonatal period were identified and compared bilirubin ¼ 460 mmol l s.d. 69.7) using the non-parametric with those without neurological abnormalities. Acute neurological test with P<0.0001. abnormalities were defined as infants having a clinical syndrome In all, 13 (41%) of the 32 infants with neurological findings consistent with acute bilirubin encephalopathy (early, intermediate had an identifiable etiology established for the hyperbilirubinemia or advanced).11,12 These include seizures, hypertonia, hypotonia, (Table 2). A total of 80 (35%) of the 226 infants without irritability, apnea, retrocollis, opisthotonus and poor suck.11,12 neurological findings had a diagnosis established for the Data on demographic characteristics, clinical course and hyperbilirubinemia. Exchange transfusions were performed in management were collected on a standardized data collection 14 (44%) of the 32 neonates with neurological abnormalities and form.10 Data were summarized using descriptive statistics for 43 (19%) of the 226 infants without abnormalities, respectively. All baseline characteristics. To compare babies with and without were treated with phototherapy. abnormalities and normal infants, the w2 test was used and the Univariate logistic regression analysis identified that infants in À1 non-parametric, Wilcoxon test was used for continuous variables. the highest peak bilirubin level group (>550 mmol l ) had the Univariate logistic regression analysis was performed to evaluate greatest risk of acute neurological abnormalities (9/27 or 30%). À1 the effect of predictors (age of infant presentation, peak serum The mid range (451 to 550 mmol l ) group had a lower rate of bilirubin, need for readmission, etiology of hyperbilirubinemia and neurological abnormalities (9/132 or 6.8%) OR of 0.174 À1 need for exchange transfusion) on neurological abnormities. The (P ¼ 0.0013). The lowest bilirubin level group (p450 mmol l ) final results are reported as OR along with their 95% confidence had 15 babies who were abnormal (15/99 or 15%) with an OR intervals. Owing to missing data in the predictors and the low of 0.402 (P ¼ 0.0613). Need for exchange transfusion and number of cases (N ¼ 32), we did not attempt to conduct presentation at p2 days of age was also associated with a multivariate analyses. The age of presentation was categorized as higher risk of acute neurological findings (Table 3). p2 days of age and >2 days of age. This time point was chosen as those with severe hyperbilirubinemia presenting early are likely to have hemolytic disease. Peak serum bilirubin was categorized as: À1 À1 À1 p450 mmol l , 451 to 550 mmol l and >550 mmol l ,as Table 1 Baseline characteristics of infants with and without abnormal these levels of hyperbilirubinemia have previously been linked to neurological findings abnormal outcomes. The significance level was set at P<0.05. a Statistical analysis was performed using the statistical software Characteristics Neurologically Neurologically P-value abnormal normal package SAS version 9.2 (SAS Institute, Cary, NC, USA). infants infants (N ¼ 32) (N ¼ 226) Results Male (%) 22 (68.8) 140 (61.9) 0.46 The initial cohort consisted of data on 258 infants with severe