University of Groningen

The management of hyperbilirubinemia in preterm infantsVader-van Imhoff, Deirdre Elisabeth

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2013

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Vader-van Imhoff, D. E. (2013). The management of hyperbilirubinemia in preterm infants. [s.n.].

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 11-04-2021

234567

9SD

8

The Management of Hyperbilirubinemia in

Preterm Infants

Deirdre E. Vader-van Imhoff

The work described in this thesis was performed within the Beatrix Childrens’ Hospital of the University Medical Center Groningen, the Netherlands. The work was financially supported by ZonMw.

The printing of this thesis was financially supported by: Rijksuniversiteit Groningen, University Medical Center Groningen, postgraduate school for Behavioral and Cognitive Neurosciences, Nutricia Nederland B.V, Nestlé Nutrition, Abbott Diagnostics and Dräger Medical Nederland B.V.

The Management of Hyperbilirubinemia in Preterm Infants © 2013 D.E. van Imhoff, the Netherlands

ISBN 978-90-367-6034-8 978-90-367-6035-5 (electronic version)

Cover design Janyte C. Holwerda Layout LINE UP boek en media, Groningen, the Netherlands

All rights reserved. No part of this thesis may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without the permission of the author or, when appropriate, of the publishers of the publications.

RIJKSUNIVERSITEIT GRONINGEN

The Management of Hyperbilirubinemia in Preterm Infants

Proefschrift

ter verkrijging van het doctoraat in de Medische Wetenschappen

aan de Rijksuniversiteit Groningen op gezag van de

Rector Magnificus, dr. E. Sterken, in het openbaar te verdedigen op

maandag 4 maart 2013 om 16:15 uur

door

Deirdre Elisabeth Vader - van Imhoffgeboren op 16 januari 1981

te Groningen

Promotor: Prof. dr. A.F. Bos Copromotores: Dr. P.H. Dijk Dr. C.V. Hulzebos Beoordelingscommissie: Prof. dr. W.P. Fetter Prof. dr. F.J. Walther Prof. dr. H.J. Verkade

 Voor mijn Vader(tjes)

Paranimfen: Janneke Licht - Hofman Karin Miedema

1234567

9SD

8

7

Table of contents

Chapter 1

Introduction 9

Part 1

The bilirubin/albumin ratio in the management of hyperbilirubinemia in preterm infants 17

Chapter 2

Kernicterus, bilirubin induced neurological dysfunction and new treatments for unconjugated hyperbilirubinemia 19 G. Buonocore et al. (eds.), Neonatology. A Practical Approach to Neonatal Diseases.

Springer-Verlag Italia 2011. Ch83:621-628

Chapter 3

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity in premature infants 37 Archives of Disease in Childhood Fetal & Neonatal Edition 2008;93:F384–F388

Chapter 4

A double-blind, randomized controlled trial on the bilirubin/albumin ratio in jaundiced preterm infants 53 Submitted

Part 2 Recommendations for the management of hyperbilirubinemia in preterm infants 75

Chapter 5 Transcutaneous bilirubin measurements in preterm infants: Effects of phototherapy and treatment thresholds 77 Submitted

8

Chapter 6

Uniform treatment thresholds for hyperbilirubinemia in preterm infants: background and synopsis of a national guideline 97 Early Human Development 2011;87:521-5

Chapter 7

High variability and low irradiance of phototherapy devices in Dutch NICUs 113 Arch Dis Child Fetal Neonatal Ed. 2012 May 18. (Epub ahead of print)

Chapter 8

Measurements of neonatal bilirubin and albumin concentrations: a need for improvement and quality control 129 European Journal of Pediatrics. 2011;170:977-82.

Chapter 9

General discussion 141

Summary 161Nederlandse samenvatting 165Dankwoord 169About the author 173List of publications 175

234567

9SD

8

1

9

Chapter I

Introduction

10

Bilirubin metabolism

Bilirubin is derived from the Latin words ‘bilis’ and ‘ruber’. An excess amount of un-conjugated bilirubin (UCB) in the blood results in yellow discoloration of the skin, i.e. jaundice, which can be observed in neonatal unconjugated hyperbilirubinemia.

Bilirubin is produced by red blood cell degradation. Red blood cells are enriched with hemoglobin which consists of four heme complexes. When heme is degraded by heme oxygenase and biliverdin reductase, water insoluble unconjugated bilirubin is formed (Figure 1). In the circulation, about 90% of the unconjugated bilirubin is bound to albumin and forms water soluble unconjugated bilirubin-albumin complexes. These unconjugated bilirubin-albumin complexes are transported to the liver were the unconjugated bilirubin is conjugated by uridine diphosphoglucuronosyltrans-ferase (UGT). Water soluble conjugated- and some unconjugated bilirubin is then transported via the biliary tree to the intestine. In conditions of severe unconjugated hyperbilirubinemia, transmucosal diffusion of UCB from the blood to the intestinal lumen is possible. In the intestinal lumen, conjugated bilirubin is hydrolyzed to un-conjugated bilirubin. Bacteria then convert the majority of the unconjugated bilirubin to urobilinogen. Urobilinogen is easily eliminated in urine and stool. Some of the unconjugated bilirubin in the intestinal lumen is reabsorbed to the circulation. This reabsorption process of unconjugated bilirubin from the intestine into the circula-tion is called the enterohepatic circulation (Figure 1).About 0.1% of the UCB is not bound to albumin and is called free bilirubin or Bf. Albumin normally has one binding site with high affinity, and several binding sites with lower affinity for UCB. Conditions associated with low albumin concentrations, or with displacement of bilirubin from albumin (e.g., by ree fatty acids or drugs such as sulfonamides) and individual variability in the binding affinity of the albumin for bilirubin may increase Bf concentrations.(1–4) Bf, and not UCB, is able to cross the blood brain barrier and may cause bilirubin neurotoxicity.(5,6)

Neonatal hyperbilirubinemia

In the 19th century it was already known that unconjugated hyperbilirubinemia could potentially harm the central nervous system of jaundiced newborn infants. Yellow staining of deep brain nuclei in jaundiced infants was first reported in 1847. The term kernicterus (in German, kern = nucleus; in Greek, ikterus = yellow) was first denoted in 1903 to describe the pathological findings of this specific yellow staining pattern.(7) Later on, kernicterus was also used to describe the neurological findings in survivors of severe hyperbilirubinemia.(7) The clinical spectrum of classical kernicterus is a well described pathological and clinical condition including an acute and chronic

11

phase of bilirubin encephalopathy. The clinical conditions associated with less severe hyperbilirubinemia are referred to as bilirubin-induced neurological dysfunction (BIND).(8,9) BIND refers to more or less subtle neurodevelopmental disabilities in infants with a history of excessive hyperbilirubinemia and signs of acute bilirubin encephalopathy. These infants can present with auditory dysfunction and/or mild neurologic abnormalities such as mild impairment in neurologic and/or cognitive performance.(8,9,10)

In the past century, kernicterus was almost exclusively seen in the context of high levels of TSB resulting from blood group incompatibility, i.e. rhesus (Rh) hemolytic disease. Rh disease became exceptional after the introduction of Rh-immune globulin. This and the introduction of phototherapy in 1970 resulted in a decrease in the inci-dence of kernicterus.(11) Recent data show remarkable differences in the incidence of kernicterus, from 1: 40.000 to 1: 150.000 live births.(12) Although infrequently oc-curring, the devastating neurological sequelae of kernicterus remain a serious threat for jaundiced newborn infants.(13,14,15) This is supported by recently published case reports on kernicterus in extreme low birth weight infants with only moderate neonatal hyperbilirubinemia.(16) Preterm infants are more prone to neurological impairment as well as bilirubin neurotoxicity than their term counterparts. 

Figure 1. Bilirubin metabolism UCB unconjugated bilirubin, Bf free bilirubin, CB conjugated bilirubin, EHC enterohepatic circulation. Adapted from chapter 2.

12

Treatment of hyperbilirubinemia is world wide based on total serum bilirubin (TSB) levels, although free bilirubin (Bf) seems a better predictor of the risk on bilirubin neurotoxicity than TSB.(5) Unfortunately, Bf measurements are not rou-tinely available. However, the TSB/Albumin (B/A) ratio has been suggested as an approximate parameter of Bf: in case of low albumin levels more Bf is available to cross the blood brain barrier. Current management guidelines for near term infants (> 35 weeks of gestational age) consider low albumin levels as risk factors for the development of bilirubin neurotoxicity. Although the B/A ratio may be also valu-able in the risk assessment of neonates with hyperbilirubinemia, the B/A ratio is not incorporated into any guideline for the management of hyperbilirubinemia.(5,17–20)

Aim of this thesis

This thesis provides recommendations aimed to improve management of neonatal hyperbilirubinemia in preterm infants. The primary aim of this thesis is to provide data on the use of the bilirubin/albumin ratio in the treatment of hyperbilirubinemia in preterm infants of less than 32 weeks of gestational age. Furthermore, several diagnostic and therapeutic processes in current care of preterm infants with hyperbilirubinemia in the Dutch neonatal intensive care units (NICUs) are analyzed.

Outline

Part 1 of this thesis focuses on neonatal hyperbilirubinemia in general and, more specifically, on the additional use of the bilirubin/albumin ratio in the management of hyperbilirubinemia in preterm infants. Chapter 2 describes the pathophysiology of bilirubin neurotoxicity, its clinical spectrum and potential diagnostic tools as well as novel treatment modalities to prevent infants from developing severe unconjugated hyperbilirubinemia and bilirubin neurotoxicity. Numerous data underline the role of free bilirubin (Bf) herein. As mentioned previously, Bf is more closely related to the development of bilirubin neurotoxicity compared to TSB.(21) To approximate the free bilirubin level, the use of the bilirubin/albumin ratio (B/A ratio) has been suggested. Chapter 3 reviews the literature on the use of the B/A ratio in predicting BIND including neurodevelopmental impairment in preterm infants of less than 32 weeks of gestational age with unconjugated hyperbilirubinemia. This resulted in a randomized controlled trial aimed to assess the concurrent use of the B/A ratio and the TSB level in the management of preterm infants with hyperbilirubinemia the B/A Ratio Trial (BARTrial) presented in Chapter 4.

13

Part 2 of this thesis is composed of 4 studies aimed to improve the management of hyperbilirubinemia in preterm infants resulting in pratical recommendations.

As previously mentioned, treatment thresholds for hyperbilirubinemia are based on the TSB level. Frequent blood sampling for TSB measurement is necessary for appropriate management of hyperbilirubinemia in preterm infants. The use of non-invasive transcutaneous bilirubin (TcB) measurements approximating total serum bilirubin levels in the management of hyperbilirubinemia is recommended in term infants.(22) Recommendations on the management of hyperbilirubinemia in preterm infants do not yet include TcB measurements. There is a paucity of evidence based data on the use of TcB measurements in preterm infants, especially when receiving phototherapy. Chapter 5 focuses on the clinical applicability of TcB measurements in preterm infants receiving phototherapy. Furthermore we analyzed effects of specific TcB cut-off levels regarding the reduction in the need for blood samples without missing severe hyperbilirubinemia.

To date evidence based treatment thresholds for hyperbilirubinemia in preterm infants are still lacking and various guidelines are used.(23) Chapter 6 explores the applied TSB treatment thresholds in the Dutch neonatal intensive care units (NICUs) and presents novel, consensus-based TSB treatment thresholds for hyperbilirubinemia in preterm infants of less than 35 weeks of GA. This guideline includes TSB-based thresholds for PT and ET for normal and high risk preterm infants.

Phototherapy is considered an efficient treatment in the majority of preterm infants with hyperbilirubinemia. Efficacy of phototherapy depends, amongst other factors, on irradiance levels. Recommendations on effective phototherapy include irradiance levels of phototherapy devices between 8 to 10 µW/cm2/nm.(22,25) In chapter 7 the clinical practice conditions of phototherapy (PT) and irradiance levels of PT devices in all Dutch NICUS are analyzed. Subsequently, recommendations are given to improve the efficacy of PT.

Obviously, variability in laboratory bilirubin (and albumin) measurements af-fects treatment decisions (start or intensify phototherapy, or perform an exchange transfusion). Data on variability of bilirubin and albumin measurements in the neonatal range are scarce and unavailable for Dutch laboratories. Chapter 8 analyzes the interlaboratory variability of neonatal bilirubin and albumin measurements in the Dutch NICUs and describes the implementation of a quality assessment scheme for these measurements.

Chapter 9 summarizes the main results of all studies included in this thesis and future perspectives are discussed to further improve the management of hyperbili-rubinemia in preterm infants.

14

References

1. Odell GB. Studies in kernicterus. I. The protein binding of bilirubin. J Clin Invest 1959 May;38(5):823–833.

2. Brodersen R. Competitive binding of bilirubin and drugs to human serum albumin studied by enzymatic oxidation. J Clin Invest 1974 Dec;54(6):1353–1364.

3. Brodersen R. Binding of bilirubin to albumin. CRC Crit Rev Clin Lab Sci 1980 Jan;11(4):305–399.

4. Odell GB, Cukier JO, Ostrea EM,Jr, Maglalang AC, Poland RL. The influence of fatty acids on the binding of bilirubin to albumin. J Lab Clin Med 1977 Feb;89(2):295–307.

5. Wennberg R, Ahlfors C, Bhutani V, Johnson L, Shapiro S. Toward Understanding Kernicterus: A Challenge to Improve the Management of Jaundiced Newborns. Pediatrics 2006 02/01;117(2):474–485.

6. McDonagh A, Maisels MJ. Bilirubin Unbound: Deja Vu All Over Again? Pediatrics 2006 02/01;117(2):523–525.

7. Hansen TW. Pioneers in the scientific study of neonatal jaundice and kernicterus. Pediatrics 2000 08;106(2):E15.

8. Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 2005 01;25(1):54–59.

9. Shapiro SM. Chronic bilirubin encephalopathy: diagnosis and outcome. Semin Fetal Neonatal Med 2010 Jun;15(3):157–163.

10. Oh W, Tyson JE, Fanaroff AA, Vohr BR, Perritt R, Stoll BJ, et al. Association Between Peak Serum Bilirubin and Neurodevelopmental Outcomes in Extremely Low Birth Weight Infants. Pediatrics 2003 10/01;112(4):773–779.

11. Dobbs RH, Cremer RJ. Phototherapy. Arch Dis Child 1975 11;50(11):833–836.

12. Maisels MJ. Neonatal hyperbilirubinemia and kernicterus - Not gone but sometimes forgotten. Early Hum Dev 2009 10/14.

13. Sugama S, Soeda A, Eto Y. Magnetic resonance imaging in three children with kernicterus. Pediatric Neurology 2001 10;25(4):328–331.

14. Watchko JF, Oski FA. Kernicterus in preterm newborns: past, present, and future. Pediatrics 1992 11;90(5):707–715.

15. Watchko JF, Jeffrey Maisels M. Enduring controversies in the management of hyperbilirubinemia in preterm neonates. Semin Fetal Neonatal Med 2010 Jun;15(3):136–140.

16. Moll M, Goelz R, Naegele T, Wilke M, Poets CF. Are recommended phototherapy thresholds safe enough for extremely low birth weight (ELBW) infants? A report on 2 ELBW infants with kernicterus despite only moderate hyperbilirubinemia. Neonatology 2011;99(2):90–94.

17. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 11;88(6):F459-F463.

18. Ahlfors CE, Parker AE. Evaluation of a model for brain bilirubin uptake in jaundiced newborns. Pediatr Res 2005 12;58(6):1175–1179.

15

19. Ahlfors CE, Wennberg RP. Bilirubin-albumin binding and neonatal jaundice. Semin Perinatol 2004 10;28(5):334–339.

20. National Institute for Health and Clinical Excellence. National Institute for Health and Clinical Excellence. Neonatal Jaundice (Clinical guideline 98). www.nice.org.uk/CG98. 2010.

21. Calligaris SD, Bellarosa C, Giraudi P, Wennberg RP, Ostrow JD, Tiribelli C. Cytotoxicity is predicted by unbound and not total bilirubin concentration. Pediatr Res 2007 11;62(5):576–580.

22. American Academy oP. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004 07;114(1):297–316.

23. Maisels MJ, Watchko JF, Bhutani VK, Stevenson DK. An approach to the management of hyperbilirubinemia in the preterm infant less than 35 weeks of gestation. J Perinatol 2012 Sep;32(9):660–664.

24. Dijk PH, de Vries TW, de Beer JJ. Guideline ‘Prevention, diagnosis and treatment of hyperbilirubinemia in the neonate with a gestational age of 35 or more weeks’. Ned Tijdschr Geneeskd 2009;153.

25. Maisels MJ, McDonagh AF. Phototherapy for neonatal jaundice. N Engl J Med 2008 02/28;358(9):920–928.

567

9SD

8

17

Part I

The bilirubin/albumin ratio in the

management of hyperbilirubinemia in preterm infants

Chapter 2

Kernicterus, bilirubin induced neurological dysfunction and new treatments for unconjugated hyperbilirubinemia 19 G. Buonocore et al. (eds.), Neonatology. A Practical Approach to Neonatal

Diseases. Springer-Verlag Italia 2011. Ch83:621-628

Chapter 3

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity in premature infants 37 Archives of Disease in Childhood Fetal & Neonatal Edition 2008;93:F384–F388

Chapter 4

A double-blind, randomized controlled trial on the bilirubin/albumin ratio in jaundiced preterm infants 53 Submitted

567

9SD

2

8

19

Chapter 2

Kernicterus, bilirubin induced neurological

dysfunction and new treatments

for unconjugated hyperbilirubinemia

Deirdre E. van Imhoff, Frans Cuperus, Peter H. Dijk, Claudio Tiribelli, Christian V. Hulzebos

G. Buonocore et al. (eds.), Neonatology. A Practical Approach to Neonatal Diseases Ch83:621–628, © Springer-Verlag Italia 2011

20

Introduction

In the 19th century it was already known that unconjugated hyperbilirubinemia could potentially harm the central nervous system of jaundiced newborn infants. Yellow staining of deep brain nuclei in jaundiced infants was first reported in 1847.

The term kernicterus (in German, kern = nucleus; in Greek, ikterus = yellow) was first denoted in 1903 to describe the pathological findings of this specific yellow staining pattern.(1) Nowadays, kernicterus is not only used to describe the patho-logical findings, but also to describe the clinical findings of acute and/or chronic bilirubin encephalopathy in jaundiced infants.(2, 3) Although acute kernicterus is an unambiguous clinical disorder in severely jaundiced newborn infants with the possibility of permanent sequelae, subtle forms of bilirubin encephalopathy referred to as bilirubin-induced neurological dysfunction, also known as BIND have evolved more recently.(4) This chapter aims to describe the pathophysiology of bilirubin neurotoxicity, its clinical spectrum and diagnostic tools. Novel treatment modalities to prevent infants from developing severe unconjugated hyperbilirubinemia and bilirubin neurotoxicity will be highlighted.

Pathophysiology – Risk Factors

The risk of kernicterus and BIND may be in part determined by the concentration of Total Serum Bilirubin (TSB), which in neonates consists almost exclusively of unconjugated bilirubin (UCB), but is primarily determined by the concentration of non-albumin bound free bilirubin (Bf). Bf can easily pass the blood-brain barrier, and may better reflect the bilirubin load distributed in the brain.(5) Several cellular mechanisms of the brain protect the brain against bilirubin accumulation. One of these protective regulating mechanisms is UCB export from the brain to the blood by multidrug resistance P-glycoprotein 1 (MDR1) and most importantly by the multidrug resistance-associated protein 1 (MRP1).(5, 6) Data from in vitro and in vivo experiments have suggested that Bf, and not TSB, is the principal determinant of bilirubin neurotoxicity.(7–9) A number of pathophysiological factors are related to the variable clinical spectrum of bilirubin neurotoxicity. First, several mechanisms including necrosis and apoptosis are involved in bilirubin-induced neuronal damage, resulting in specific types of cellular damage and dysfunction.(5, 6, 10) Second, neural susceptibility to bilirubin neurotoxicity is not comparable in all cell types of the brain, neurons being more susceptible than glia cells with the exact mechanisms involved remaining speculative.(5, 6, 10) Third, the amount of UCB that enters the brain is dependent not only on the Bf concentration, but also on the intactness of the blood

Kernicterus, bilirubin induced neurological dysfunction and new treatments 21

brain barrier. The Bf concentration in plasma is determined by the concentration of unconjugated bilirubin bound to albumin, at a specific plasma pH, in relation to the affinity of bilirubin to bind albumin. Conditions associated with low albumin con-centrations, or with displacement of bilirubin from albumin (e.g., by sulfonamides or free fatty acids) increase Bf concentrations. Conditions that decrease intactness of the blood-brain barrier (e.g., hyperosmolality, hypercapnia, asphyxia, prematurity, infection, and sepsis) can result in a net increase of UCB uptake in the brain.(4) Multiple other factors (i.e., cerebral blood flow, vascular permeability, cellular efflux pumps and cellular recovery capacity) may affect the development of neurotoxic-ity at any given Bf concentration.(6) Some of these factors are incorporated as risk factors in management guidelines of jaundiced infants. Clinical evidence of most of these risk factors is limited. Most factors are based on anecdotal clinical evidence or theoretical and experimental animal data. Although imperfect, these factors are used by many clinicians to determine TSB thresholds before starting treatment.(2, 3, 11–15) The risk factors that increase the susceptibility for bilirubin neurotoxicity and risk factors that pre-dispose to develop hyperbilirubinemia are shown in Table 1. Risk assessment of bilirubin neurotoxicity for each individual newborn infant involves more than measurement of the TSB concentration, and appears to be a dynamic process affected by continuously changing individual conditions. Unfortunately, Bf cannot be routinely measured so that concentration of TSB and the currently used risk factors are the principle parameters that determine treatment thresholds.

Table 1. Risk factors used in addition to TSB to assess the risk of bilirubin neurotoxicity in the management of jaundiced infants

Risk Factors

Acidosis

Asphyxia

Hemolysis

Hypothermia

Hypoalbuminemia or less albumin available to bind bilirubin

Intracranial hemorrhage

Low birth weight

Meningitis

Prematurity

Sepsis

22

Clinical Symptoms

The clinical spectrum of signs of bilirubin neurotoxicity relates to the bilirubin-induced damage to specific brain areas. Specific brain stem nuclei (auditory, vestibular and oculomotor), cerebellar Purkinje cells, basal ganglia (i.e., globus pallidus and subthala-mus) and the hippocampus are particularly vulnerable to bilirubin neurotoxicity.(6, 16) In the initial phase, acute bilirubin encephalopathy is characterized by lethargy, hypotonia and poor sucking. In the second or intermediate phase, most commonly within 2–3 days after the initial phase, hypertonia (retrocollis, opistotonus), discomfort and irritability dominate. The infant may develop fever, a typical high-pitched cry, and seizures, which may alternate with drowsiness and hypotonia. Exchange transfusion at this stage might reverse the central nervous system damage in some cases. When left untreated, this second phase progresses into a more advanced stage characterized by pronounced retrocollis-opisthotonus, shrill cry, apnea, fever, sometimes seizures, deep stupor to coma and even death.(3,17) Choreoathethosis, vertical gaze paralysis, auditory dysfunction, and dental dysplasia (Perlstein’s tetrad) and motor delay are the most commonly described permanent neurologic sequelae in infants who survive the acute phase of kernicterus. Intelligence is usually normal in these infants.(5,17) Less severe hyperbilirubinemia can result in subtle bilirubin encephalopathy referred to as ‘bilirubin- induced neurological dysfunction’ (BIND) or as ‘bilirubin- associated neurological dysfunction’ (BAND).(18) Subtle permanent bilirubin encephalopathy can present with auditory dysfunction and/or mild neurologic abnormalities such as mild impairment in neurologic and/or cognitive performance.(4,19) Involvement of the auditory nervous system dysfunction may result in sensorineuronal hearing loss or deafness. Alternatively, auditory dysfunction referred to as auditory neuropathy (AN) or auditory dys-synchrony (AD) may occur. AN/AD is defined as a normal neurophysiological test of the inner ear, i.e., normal cochlear microphonic responses and otoacoustic emissions, but an abnormal or absent auditory brainstem response (ABR) resulting in abnormal sound processing. AN/AD is clinically characterized by problems in sound localization and speech discrimination. Hearing loss is not present per se and a normal audiogram is often seen.(4)

Several studies have evaluated effects of moderate and severe degees of hyper-bilirubinemia with respect to neurodevelopmental outcome in later childhood.(20) Considering the differences in incidence and severity of neurologic dysfunction in infants with a comparable severity of neonatal hyperbilirubinemia, it seems impos-sible to predict outcome by TSB levels in jaundiced newborn infants. In general, preterm infants and infants suffering hemolytic disease seem to have an increased risk of neurologic sequelae of neonatal hyperbilirubinemia.(20)

Kernicterus, bilirubin induced neurological dysfunction and new treatments 23

Isolated movement disorders such as athetosis or distonia are sometimes seen and in a retrospective analysis of extremely low birth weight infants a significant as-sociation between peak concentrations of TSB during the first two weeks of life and neurodevelopmental impairment was found.(4, 17, 19, 21, 22)

Epidemiology

In the past century, kernicterus was almost exclusively seen in the context of high concentrations of TSB related to Rh hemolytic disease. When Rh disease became rare due to the introduction of Rh-immune globulin, and phototherapy appeared an effective treatment for unconjugated hyperbilirubinemia, the incidence of kernicterus decreased. Subsequently, treatment criteria were liberalized. As a combined result of the more liberal treatment criteria, a trend towards earlier hospital discharge (before the maximal TSB concentration is reached), and higher survival rate of preterm in-fants, kernicterus remains a serious threat for jaundiced newborn infants. Nowadays, the exact incidence of kernicterus and BIND is unknown. This is, at least in part, related to differences in the definition of severe hyperbilirubinemia and in methods of assessing neurodevelopmental outcome among studies. In a prospective study investigating severe hyperbilirubinemia (defined as maximum TSB >510 µmol/L) of the newborn in the UK, the incidence of bilirubin encephalopathy was ~1 case per 100,000 live births.(23) It is estimated that 1–3 per 100,000 live births are at risk of developing kernicterus when untreated and that 5–10% of infants surviving severe hyperbilirubinemia suffer permanent sequelae.(6, 20)

Diagnosis

Current management guidelines for jaundiced infants are actually based on the total serum bilirubin (TSB) concentrations. Because exact neurotoxic TSB concentrations are unknown and risk factors for imminent bilirubin neurotoxicity are not evidence based, other diagnostic tools, in addition to TSB, may be valuable to detect imminent BIND in jaundiced newborn infants (Table 2).

To evaluate the risk of BIND, the history of the infant is important. Information about the gestational age, physical examination and the presence of risk factors for bilirubin neurotoxicity as well as the duration of the hyperbilirubinemic period, the presence of acute symptoms and whether the child has previously been treated or not is crucial data to diagnose BIND at a later age.(24)

Although preliminary clinical series and in vitro data point to a correlation between Bf and BIND, the difficulty of clinically measuring Bf has prevented its introduction

24

in clinical practice. Alternatively, the Bilirubin/Albumin ratio (B/A ratio) could be used as a surrogate parameter to indicate Bf concentrations.(25,26) However, due to the presence of bilirubin displacers, i.e., drugs that interfere with the bilirubin- albu-min binding, Bf may reach values much higher than suggested by the calculated B/A ratio. In addition, the individual variability in the binding affinity of the albumin for bilirubin should also be taken into account. Alternatively, other plasma constituents such as apolipoproteins may bind UCB, and decrease Bf. Consequently, the B/A ratio seems an imperfect surrogate to estimate Bf. It needs to be outlined however, that Bf may also be an imperfect predictor of bilirubin neurotoxicity, since many factors affect the development of bilirubin neurotoxicity at any given Bf level.

The theoretical considerations and clinical evidence for the concept that additional use of the B/A ratio, i.e., next to TSB, in jaundiced premature infants might improve the prediction of BIND has been recently reviewed. Although no prospective clinical trials exist, it is suggested that the additional use of the B/A ratio may be valuable in evaluating jaundiced premature infants.(27–31) A randomized controlled trial inves-tigating the additional use of the B/A ratio in the treatment of hyperbilirubinemia in preterm infants (ISRCTN74465643) is underway and will hopefully provide the answer to this important question.

The possibility that biochemical markers, i.e., use of Tau and S100B protein con-centrations, are useful in the diagnosis of bilirubin-induced neurotoxicity has been studied in hyperbilirubinemic newborn infants.(32) Tau is a microtubule-associated structural protein of central nervous system neurons. S100B protein, a neurotrophic factor, is synthesized in astroglial cells in the central nervous system and Schwann cells. Previous data showed increased Tau and S100B protein concentration in plasma and/or cerebrospinal fluid in patients with cerebral injury (e.g., hypoxia and trauma). In a prospective study of 92 hyperbilirubinemic non-asphyxiated infants, Tau and S100B protein concentration was positively correlated with TSB concentration. Tau and S100B concentration increased at TSB concentration above 327 µmol/L, as observed in 46 infants. At lower TSB concentrations, protein concentrations re-mained unchanged. Above a TSB concentration of 327 µmol/L, clinical symptoms of bilirubin encephalopathy (i.e., auditory neuropathy, minor neurologic dysfunction and electroencephalographic abnormalities) were present in 22 of 46 of the infants. Compared to TSB concentrations, measurement of Tau and S100B protein concen-tration did not improve sensitivity or specificity for any of the described clinical symptoms. The use of Tau and S100B protein concentration in the assessment of BIND is not recommended.

The neural auditory pathway is very susceptible to bilirubin toxicity putatively re-sulting in sensorineurinal hearing loss or auditory neuropathy also known as auditory

Kernicterus, bilirubin induced neurological dysfunction and new treatments 25

dys-synchrony. A frequently used, non-invasive and very sensitive tool to determine bilirubin neurotoxicity, is the Auditory Brainstem Response (ABR), which allows for determination of electrophysiological activity of the neural auditory pathway. The ABR consists of a sequence of positive waves (numbered I–V) representing the auditory pathway from inner ear to brainstem. Wave I and II represent the peripheral auditory nerve and waves III–V represent the activity in the auditory centers at the brain stem level of the pathway (cochlear nucleus and lateral lemniscus, respectively).(33) Bilirubin-induced ABR changes mainly involve waves III and V and may progress from reversible increased interwave latencies to the loss of wave amplitude. ABR changes can be transient, but may also progress into permanent wave changes or even loss of any recognizable wave.(7,27) A bedside method to evaluate the intactness of the auditory pathway is the Automated Auditory Brainstem Response (AABR) with an ALGO hearing screening system (Natus Medical, San Carlos, CA, USA). AABR measurements are simplified ABR measurements and able to identify infants with abnormal cochlear or auditory function. A pass or refer result is shown on the ALGO machine for each ear of the infant. In an observational study of 83 Kernicterus, Bilirubin Induced Neurological Dysfunction and New Treatments for Unconjugated Hyperbilirubinemia 191 patients of variable birth weight (406–4727 g) and variable gestational age (24–42 weeks), an abnormal ALGO result (bi-or unilateral refer) was associated with increased Bf concentrations and Bf /TSB ratios, but not with TSB concentrations alone.(34)

Another tool to identify acute and chronic bilirubin neurotoxicity in jaundiced infants is Magnetic Resonance Imaging (MRI).(18) MRI changes include bilateral hyperintensity of globus pallidus on T1-weighted scans in the early phase (first 3 weeks of life) and most commonly subtle but persis - tent pallidal hyperintensity, suggestive of permanent gliosis, on late T2-weighted scans (with disappearance of hyperintensity on T1-weighted scans). The subthalamic nucleus and hippocampus are affected less frequently (approximately 40% and 5% of reported cases, respectively). Bilateral dam-age of the globus pallidus and the subthalamic nucleus are specific signs of bilirubin neurotoxicity, which sometimes can also be visualized with ultrasonography, and is a key in the differentiation between hypoxic-ischemic encephalopathy or metabolic disorders, which predominantly affect the thalamus.(30) The additional value of MR spectroscopy to MRI is not completely clear, but acute changes in cerebral metabolism have been shown in bilirubin encephalopathy.(35)

In 1999, a clinical scoring system was developed to evaluate the risk of exposure to unconjugated hyperbilirubinemia. Clinical signs of bilirubin encephalopathy include mental status, muscle tone and cry of the infant. Dependent on the level of abnormality, 0–3 points are obtained for each clinical sign, resulting in an overall

26

score between 0 and 9, representing no toxicity or advanced toxicity, respectively. Validation of this scoring system is in progress.(36, 37)

Table 2. Diagnostic tools to evaluate the severity of hyperbilirubinemia and/or risk of imminent bilirubin neurotoxicity

Diagnostic tool

Anamnestic History of hyperbilirubinemic periodTreatment for neonatal hyperbilirubinemiaGestational ageRisk factors present in hyperbilirubinemic periodSymptoms of kernicterus during hyperbilirubinemic periodAbnormal physical examination

Total serum bilirubin concentration

Free bilirubin concentration

Bilirubin/albumin ratio

Auditory brainstem response

Magnetic resonance imaging

BIND score

BIND = Bilirubin induced neurological dysfunction

Treatment

To identify newborns at risk for severe hyperbilirubinemia and to prevent bilirubin neurotoxicity, guidelines have been developed for the management of jaundiced in-fants. The American Academy of Pediatrics’ Subcommittee on Hyperbilirubinemia adapted a guideline for hyperbilirubinemic infants of 35 or more weeks of gestation in 2004.(3) Currently, this guideline is adapted by many countries worldwide for the management of jaundiced ‘near term’ and term newborn infants. However, international guidelines for preterm infants are lacking and rather local or national guidelines for preterm infants are used, such as the consensus-based guideline on hyperbilirubinemia for preterm infants of 35 or less weeks of gestational age in The Netherlands (www.neonatologiestudies.nl/main/richtlijnen).(38) Few prospec-tive studies have analyzed effects of different TSB-thresholds on longterm out-come (Table 3). In a group of 95 low birth weight infants, hyperbilirubinemia was treated either with prophylactic phototherapy (starting 12 hours after birth, n = 46) or conservative phototherapy (starting at fixed TSB threshold concentrations of 150 µmol/L, n = 49). The maximum TSB concentrations were comparable in both groups, but in a subgroup of extreme low birth weight infants (ELBW, < 1000 g), the

Kernicterus, bilirubin induced neurological dysfunction and new treatments 27

maximum TSB concentrations were significantly higher in the conservative group (171 µmol/L versus 139 µmol/L, conservative versus prophylactic treatment) with a concomitant higher incidence of cerebral palsy. However, since the study was not originally powered for this outcome, it failed to demonstrate significant differences in long-term neurodevelopmental outcome.(39) Another randomized controlled trial assigned 1974 ELBW preterm infants to a prophylactic phototherapy group (started immediately postnatal) or a conservative phototherapy group (based on predefined TSB threshold concentrations: 137 µmol/L for infants weighing 501–750 g and 171 µmol/L for infants weighing 751–1000 g). The primary outcome of this study was a combination of neurodevelopmental impairment (defined as blindness, severe hearing loss, moderate of severe cerebral palsy, or a score below 70 on the mental or psychomotor developmental index of the Bayley Scale of Infant Development II) or death. In this study, maximum TSB concentrations were significantly higher in the conservative group: 168 µmol/L versus 120 µmol/L in the prophylactic group. Prophylactic phototherapy significantly reduced the risk of neurodevelopmental impairment (relative risk (95% confidence intervals): 0.86 (0.74–0.99)). This benefit of prophylactic phototherapy seems questionable for infants with a birth weight of 501–750 g, because in this group a non-significant trend of increased mortality was found.(40)

Table 3. Long-term outcome of randomized controlled trials on TSB thresholds

Author/ Year

Population Phototherapy Long term outcome

Pro phylactic Conservative

Morris, 2008(1)

ELBW infants (BW <1000 g.) n=1974

direct post-natally

137 µmol/L (BW 501– 750 g) or 171 µmol/L (BW 751–1000g)

No significant difference for the combination of death or neuro-logical impairment Prophylactic phototherapy reduced the risk of neurodevelopmental impairment (RR=0.86 (95% CI: 0.74 to 0.99) Non-significant higher mortal-ity among infants with a BW of 501–750 g (RR=1.13 (95% CI:0.96 to 1.34)

Jangaard, 2007(2)

LBW infants (<1500 g) n=95

12 hours postnatally

150 µmol/L Non-significant increase in cere-bral palsy and death in a sub-group of infants (with a BW <1000 g) in the conservative group

ELBW: extreme low birth weight, i.e., birth weight less than 1000 grams; LBW: low birth weight, i.e., birth weight less than 1500 grams. RR: relative risk, CI: Confidence Intervals. 17.1 µmol/L = 1 mg/dL bilirubin

28

New Treatments for Unconjugated Hyperbilirubinemia

Conventional treatment for severe unconjugated hyperbilirubinemia consists of pho-totherapy and exchange transfusion. Although phototherapy is considered effective and safe, it does not always prevent toxic accumulation of bilirubin in neonates. Long-term phototherapy, as required by Crigler- Najjar patients may take up to 16 hours per day and becomes less effective with age. In spite of this intensive treatment regimen, approximately 30% of the patients with Crigler-Najjar disease type I develop mild to severe brain damage. In addition, patients still die resulting from complications related to the disease.(41) Exchange transfusion should be considered if phototherapy fails to decrease plasma bilirubin below toxic levels. This ‘rescuetreatment’ however, is associated with a significant morbidity in sick newborns, and even mortality has been reported. These considerations have prompted investigators to develop alterna-tive treatments for unconjugated hyperbilirubinemia.(42) Most of these treatments are still in an experimental phase. Nevertheless, it is highly conceivable that some of them will find their way into clinical practice in the foreseeable future.

Treatments That Decrease the Production of Bilirubin

Treatments that decrease the production of bilirubin could actually prevent uncon-jugated hyperbilirubinemia, which would be a more rational strategy compared with removing bilirubin that is already accumulated in the body (see Chapter 84). Bilirubin is produced in the macrophages of the reticuloendothelial system. These macophages contain two essential enzymes: heme oxygenase (HO) and biliverdin reductase. Heme oxygenase converts heme, the source of bilirubin, into blue-green biliverdin. Biliverdin reductase then converts biliverdin into bilirubin. Consequently, bilirubin production can be decreased by agents that inhibit HO, and/or biliverdin reductase (Fig. 1). Several agents have been developed to inhibit these enzymes such as the metalloporphyrins.(43,44) Metalloporphyrins compete with heme for HO binding sites resulting in a competitive inhibition of HO and in a decreased conversion of heme into biliverdin. Currently, tinmesoporphyrin is the most evalu-ated metalloporphyrin in humans. In the 8 clinical trials so far conducted, it was demonstrated that this agent mitigates unconjugated hyperbilirubinemia due to ABO incompatibility and prematurity. Also, the use of tin-mesoporhyrin replaced the need for phototherapy in glucose-6-phosphate dehydrogenase deficient new-borns. Tin-mesoporphyrin thus seems a highly promising new treatment strategy for unconjugated hyperbilirubinemia. Currently, however, metalloporhyrins cannot be recommended for routine treatment of unconjugated hyperbilirubinemia due to insufficient evidence regarding the long-term safety of these agents.

Kernicterus, bilirubin induced neurological dysfunction and new treatments 29

D-penicillamine, a chelating agent that is used in the treatment of Wilson’s disease, is another HO-inhibitor (Fig. 1).(43) The use of this agent decreased the number of exchange transfusions in infants with ABO-hemolytic disease, but its efficacy has only been evaluated in a single clinical trial. Curiously, inhibitors of biliverdin reductase have never been explored, most likely because the accumulation of biliverdin in the blood would produce green babies.

Figure 1. Inhibitors of heme oxygenase effectively decrease the conversion of heme into biliverdin, the first essential step in bilirubin production. Reproduced from (43), with permission

Treatments That Increase the Hepatic Clearance of Biliribin

Almost all bilirubin is excreted via the bile, and many therapies aim to enhance this excretory pathway by targeting the hepatic clearance of the bilirubin. Previously it has been described how newly produced bilirubin enters the microcirculation of the liver from where it is extracted by the hepatocytes. In the hepatocyte, bilirubin is firstly bound to cytosolic ligandin, which prevents its diffusion back into the blood. The enzyme UDP-glucuronosyltransferase 1A1 (UGT1A1), which resides in the endoplasmatic reticulum of the hepatocytes, subsequently conjugates bilirubin with one or two glucuronic acid groups. This step increases the water solubility of the molecule and allows the now conjugated bilirubin to be excreted into the bile via the canalicular transporter MRP2. Several therapies target these 3 steps in bilirubin catabolism (Fig. 2). Clofibrate, a lipid-lowering drug, increases UGT1A1 activity (Fig. 2). This drug mitigated neonatal unconjugated hyperbilirubinemia in several clinical trials, and decreased the need for phototherapy if applied as add-on treatment. However, long-term clofibrate use is associated with an overall increase in non-cardiovascular mortality. Although short-term clofibrate treatment has not been shown to have serious adverse effects, its safety issues must be addressed before it could be considered for clinical use.

30

The most well-known agent that enhances hepatic bilirubin clearance is pheno-barbital. This anti-epileptic agent enhances ligandin, UGT1A1, and MRP2 activity (Fig. 2). Since the 1960s, phenobarbital has been used extensively in the treatment of neonatal jaundice. Numerous clinical trials have shown that administering pheno-barbital to pregnant women or to newborns mitigates neonatal hyperbilirubinemia and decreases the number of exchange transfusions. Nevertheless, phenobarbital is currently not used as a routine treatment of neonatal jaundice. This is mainly because phototherapy is more effective, decline of TSB concentrations in infants treated with phenobarbital and phototherapy is not faster when compared to phototherapy alone, and because its therapeutic effect is not evident until a few days after administra-tion, in contrast to some of the adverse effects (i.e., sedation).(43) Phenobarbital is still applied in Crigler-Najjar type II patients, who have a 95% decrease in UDP-glucuronosyltransferase activity. In these patients Phenobarbital is able to increase the residual enzyme activity, which effectively counteracts the development of severe unconjugated hyperbilirubinemia. Phenobarbital is not effective in the treatment of Crigler-Najjar type I, however, because these patients lack residual (e.g., inducible) UDP-glucuronosyltransferase activity.

The most effective treatment for Crigler-Najjar disease type I would be to repair or replace the defective UGT1A1 in the liver. Currently, this can only be achieved by a liver transplantation. In 1986 the first successful orthotopic liver transplantation (OLT) in a Crigler-Najjar patient was reported, and several other patients have un-dergone transplantation since.

Two types of transplantation have been used: OLT in which the patients’ own liver is replaced by a donor liver, and auxiliary liver transplantation in which part of the own liver is left in situ and is supported by a donor graft. If successful, a liver transplant completely corrects the underlying metabolic defect, which dramatically improves the quality of life. Liver transplantation, however, remains a high-risk procedure, with a one-year survival between 85% and 90%. Infusion of hepatocytes with an unimpaired UGT1A1 activity into the liver of Crigler-Najjar patients would be an attractive alternative to a liver transplant. This procedure, which is known as hepatocyte transplantation, could partially restore enzyme activity without the many complications that are associated with a liver transplant. So far, 7 Crigler-Najjar type I patients have received this treatment, often with multiple infusions of hepatocytes (for example via the portal vein). Hepatocyte infusion, however, decreased plasma bilirubin levels only for a limited period of 5–6 months and did not eliminate the need for a liver transplant in these patients.(45,46) Ultimately, gene therapy would be the most elegant method to repair or replace defective UGT1A1 within the hepa-tocytes of Crigler-Najjar type I patients. The results of gene therapy seem promising

Kernicterus, bilirubin induced neurological dysfunction and new treatments 31

in animal models.(47) Currently, however, the results of ongoing clinical trials must be awaited before gene therapy can be applied in Crigler-Najjar patients.

Figure 2. Phenobarbital and clofibrate increase the hepatic clearance of bilirubin. Alb albumin, UCB unconjugated bilirubin, UGT1A1 UDP-glucuronosyltransferase1A1. Reproduced from (43), with permission

Treatments That Decrease the Enterohepatic Circulation of Bilirubin

After conjugation, bilirubin enters the intestinal lumen via the bile. The intestinal con-jugated bilirubin is subsequently mostly hydrolyzed to UCB, which can be reabsorbed into the enterohepatic circulation (Fig. 3a). The majority of this reabsorbed UCB, however, spills over into the systemic circulation due to the poor first pass extraction by the liver. Conditions that enhance this enterohepatic circulation contribute to the pathogenesis of unconjugated hyperbilirubinemia. For example, poor feeding in neonates is associated with increased plasma bilirubin levels most likely caused by a delayed gastrointestinal transit, which increases the amount of intestinal UCB available for reabsorption. Indeed, conditions that accelerate the gastrointestinal transit, such as early and frequent feedings, seem to lower plasma bilirubin levels in newborn infants.(48,49) Bilirubin reabsorption could also be prevented by agents that bind and capture the pigment within the intestinal lumen (Fig. 3b).

Several capturing strategies have so far been tested. Cholestyramine, a known binder of bile salts, and agar, a gelatinous substance derived form seaweed, decreased plasma bilirubin levels in hyperbilirubinemic rats, but was less effective in neonates. Treatment of neonates with activated charcoal effectively decreased plasma bilirubin

32

levels, but only if it was administered within the first day of life. Charcoal, however, might also capture essential nutrients, which limits its clinical applicability.(43) Oral amorphous calcium phosphate did lower plasma bilirubin levels, but only in Crigler-Najjar type I patients. Calcium phosphate is currently used in a number of Dutch Crigler-Najjar type I patients as an adjunct to phototherapy if plasma uncon-jugated bilirubin levels become dangerously elevated. Orlistat, a lipase inhibitor, increases the intestinal fat content, which is hypothesized to capture the lipohilic unconjugated bilirubin. Orlistat has been shown to lower plasma bilirubin levels in a trial with Crigler-Najjar patients, but the decrease was considered clinically relevant (i.e., >10%) in only 7 of 16 patients. Zinc salts are also well-known binders of uncon-jugated bilirubin, and moderately decreased plasma bilirubin levels in patients with Gilbert’s syndrome.(43) This inherited condition is characterized by a chronic, mild unconjugated hyperbilirubinemia related to diminished hepatic UGT1A1 expression. Administration of zinc salts, however, may lead to increased zinc levels in plasma, which may limit their clinical use.

Figure 3. A. Normally, and especially in conditions that delay the intestinal transit, there is an enterohepatic circulation of unconjugated bilirubin which increases plasma bilirubin levels. B. Binding unconjugated bilirubin to capture agents in the intestinal lumen may prevent this enterohepatic circulation and thus decrease plasma bilirubin levels. UCB unconjugated bilirubin

CB conjugated bilirubin, EHC enterohepatic circulation. Reproduced from (43), with permission

Kernicterus, bilirubin induced neurological dysfunction and new treatments 33

Future Prospective

As indicated above, the molecular events leading to BIND are still partially unde-fined, as the criteria are not fully described to define BIND when the damage may be successfully treatable. It is therefore necessary to approach this increasingly present damage in a more translational way linking in vitro with in vivo models. This will hopefully provide hints on an effective prevention of the neurological damage in addition to the established phototherapy. To this end we need to better understand how unconjugated, free bilirubin may enter the cell and how the cells handle this potentially toxic substance. We know that MRP1 may extrude UCB from cells but data obtained in vitro suggest that the activity of this ABC transporter is not the only player in reducing UCB cytotoxicity. Intracellular oxidation of UCB may be an additional mechanism that needs to be assessed, as pharmacological inducers of oxidizing pathways are available. Also it is still unknown why UCB accumulates and damages only certain regions of the brain. Understanding this unexplained uneven distribution will unravel the background for regional sensitivity. Collectively these data will hopefully allow the definition of more effective treatments not only for the newborns but also for patients with Crigler-Najjar type I.

34

References

1. Hansen TW (2000) Pioneers in the scientific study of neonatal jaundice and kernicterus. Pediatrics 106:E15

2. Watchko JF, Maisels MJ (2003) Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 88:F455–F458

3. American Academy of Pediatrics Subcommittee on Hyperbilirubinemia (2004) Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 114:297– 316

4. Shapiro SM (2005) Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 25:54–59 a b

5. Ostrow JD, Pascolo L, Shapiro SM, Tiribelli C (2003) New concepts in bilirubin encephalopathy. Eur J Clin Invest 33:988–997

6. Wennberg R, Ahlfors C, Bhutani V et al (2006) Toward understanding kernicterus: a challenge to improve the management of jaundiced newborns. Pediatrics 117:474–485

7. Ahlfors CE, Shapiro SM (2001) Auditory brainstem response and unbound bilirubin in jaundiced (jj) Gunn rat pups. Biol Neonate 80:158–162

8. Ostrow JD, Pascolo L, Tiribelli C (2003) Reassessment of the unbound concentrations of unconjugated bilirubin in relation to neurotoxicity in vitro. Pediatr Res 54:98–104

9. Calligaris SD, Bellarosa C, Giraudi P et al (2007) Cytotoxicity is predicted by unbound and not total bilirubin concentration. Pediatr Res 62:576–80

10. Ostrow JD, Pascolo L, Brites D, Tiribelli C (2004) Molecular basis of bilirubin-induced neurotoxicity. Trends Mol Med 10:65–70

11. Pearlman MA, Gartner LM, Lee K et al (1980) The association of kernicterus with bacterial infection in the newborn. Pediatrics 65: 26–29

12. Kim MH, Yoon JJ, Sher J, Brown AK (1980) Lack of predictive indices in kernicterus: a comparison of clinical and pathologic factors in infants with or without kernicterus. Pediatrics 66:852–858

13. Turkel SB, Guttenberg ME, Moynes DR, Hodgman JE (1980) Lack of identifiable risk factors for kernicterus. Pediatrics 66:502–506

14. Lucey JF (1972) Neonatal jaundice and phototherapy. Pediatr Clin North Am 19:827–839

15. Maisels MJ, Watchko JF (2003) Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 88:F459–F463

16. Amin SB (2004) Clinical assessment of bilirubin-induced neurotoxicity in premature infants. Semin Perinatol 28:340–347

17. Shapiro SM (2003) Bilirubin toxicity in the developing nervous system. Pediatr Neurol 29:410–421

18. Volpe JJ (2009) Bilirubin and brain injury. In: Neurology of the Newborn 5th edn. Saunders Elsevier, Philadelphia, pp 619–651

Kernicterus, bilirubin induced neurological dysfunction and new treatments 35

19. Oh W, Tyson JE, Fanaroff AA et al (2003) Association between peak serum bilirubin and neurodevelopmental outcomes in extremely low birth weight infants. Pediatrics 112:773–779

20. Ip S, Chung M, Kulig J et al (2004) An evidence-based review of important issues concerning neonatal hyperbilirubinemia. Pediatrics 114:e130–e153

21. Newman TB, Klebanoff MA (1993) Neonatal hyperbilirubinemia and long-term outcome: another look at the Collaborative Perinatal Project. Pediatrics 92:651–657

22. Grimmer I, Berger-Jones K, Buhrer C et al (1999) Late neurological sequelae of non-hemolytic hyperbilirubinemia of healthy term neonates. Acta Paediatr 88:661–663

23. Manning D, Todd P, Maxwell M, Jane PM (2007) Prospective surveillance study of severe hyperbilirubinaemia in the newborn in the UK and Ireland. Arch Dis Child Fetal Neonatal Ed 92:F342–F346

24. Shapiro SM (2010) Chronic bilirubin encephalopathy: diagnosis and outcome. Semin Fetal Neonatal Med 15:157–163

25. Ahlfors CE, Wennberg RP (2004) Bilirubin-albumin binding and neonatal jaundice. Semin Perinatol 28:334–339

26. Ahlfors CE, Parker AE (2005) Evaluation of a model for brain bilirubin uptake in jaundiced newborns. Pediatr Res 58:1175–1179

27. Amin SB, Ahlfors C, Orlando MS et al (2001) Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 107: 664–670

28. Scheidt PC, Graubard BI, Nelson KB et al (1991) Intelligence at six years in relation to neonatal bilirubin levels: follow-up of the National Institute of Child Health and Human Development Clinical Trial of Phototherapy. Pediatrics 87:797–805

29. Ritter DA, Kenny JD, Norton HJ, Rudolph AJ (1982) A prospective study of free bilirubin and other risk factors in the development of kernicterus in premature infants. Pediatrics 69:260–266

30. Govaert P, Lequin M, Swarte R et al (2003) Changes in globus pallidus with (pre)term kernicterus. Pediatrics 112(6 Part 1):1256– 1263

31. Hulzebos CV, Van Imhoff DE, Bos AF et al (2008) Usefulness of bilirubin/ albumin ratio for predicting bilirubin-induced neurotoxicity in premature infants. Arch Dis Child Fetal Neonatal Ed 93: F384–F388

32. Okumus N, Turkyilmaz C, Onal EE et al (2008) Tau and S100B proteins as biochemical markers of bilirubin-induced neurotoxicity in term neonates. Pediatr Neurol 39:245–252

33. Amin SB, Orlando MS, Dalzell LE et al (1999) Morphological changes in serial auditory brain stem responses in 24 to 32 weeks’ gestational age infants during the first week of life. Ear Hear 20: 410–418

34. Ahlfors CE, Amin SB, Parker AE (2009) Unbound bilirubin predicts abnormal automated auditory brainstem response in a diverse newborn population. J Perinatol 29:305–309

35. Groenendaal F, van der Grond J, de Vries LS (2004) Cerebral metabolism in severe neonatal hyperbilirubinemia. Pediatrics 114: 291–294

36

36. Bilirubin-induced Neurologic Dysfunction (BIND) Among Nigerian Infants. www.med.umn.edu/peds/global/research/prevalence_of_ bilirubinemia/home.html

37. Johnson L, Brown AK, Bhutani VK (1999) BIND: A clinical score for bilirubin induced neurologic dysfunction in newborns. Pediatr Suppl 104:746

38. Van Imhoff DE, Dijk PH, Hulzebos CV; on behalf of the BARTrial studygroup of the Netherlands Neonatal Research Network (2011) Uniform treatment thresholds for hyperbilirubinemia in preterm infants: background and synopsis of a national guideline. Early Hum Dev (Epub ahead of print)

39. Jangaard KA, Vincer MJ, Allen AC (2007) A randomized trial of aggressive versus conservative phototherapy for hyperbilirubinemia in infants weighing less than 1500 g: Short- and long-term outcomes. Paediatr Child Health 12:853–858

40. Morris BH, Oh W, Tyson JE et al (2008) Aggressive vs. conservative phototherapy for infants with extremely low birth weight. N Engl J Med 359:1885–1896

41. Van Der Veere CN, Sinaasappel M, McDonagh AF et al (1996) Current therapy for Crigler-Najjar syndrome type 1: report of a world registry. Hepatology 24:311–315

42. Dennery PA, Seidman DS, Stevenson DK (2001) Neonatal hyperbilirubinemia. N Engl J Med 344:581–590

43. Cuperus FJ, Hafkamp AM, Hulzebos CV, Verkade HJ (2009) Pharmacological therapies for unconjugated hyperbilirubinemia. Curr Pharm Des 15:2927–2938

44. Suresh GK, Martin CL, Soll RF (2003) Metalloporphyrins for treatment of unconjugated hyperbilirubinemia in neonates. Cochrane Database Syst Rev 2:CD004207

45. Gupta S, Chowdhary JR (1992) Hepatocyte transplantation: back to the future. Hepatology 15:156–162

46. Ito M, Nagata H, Miyakawa S, Fox IJ (2009) Review of hepatocyte transplantation. J Hepato-biliary Pancreat Surg 16:97–100

47. Miranda PS, Bosma PJ (2009) Towards liver-directed gene therapy for Crigler-Najjar syndrome. Curr Gene Ther 9:72–82

48. Wu PY, Teilmann P, Gabler M et al (1967) “Early” versus “late” feeding of low birth weight neonates: effect on serum bilirubin, blood sugar, and responses to glucagon and epinephrine tolerance tests. Pediatrics 39:733–739

49. Wennberg RP, Schwartz R, Sweet AY (1966) Early versus delayed feeding of low birth weight infants: effects on physiologic jaundice. J Pediatr 68:860–866

567

9SD

234

8

37

Chapter 3

Usefulness of the bilirubin/albumin ratio

for predicting bilirubin-induced neurotoxicity

in premature infants

Christian V. Hulzebos, Deirdre E. van Imhoff, Arend F. Bos, Charles E. Ahlfors, Henkjan J. Verkade, Peter H. Dijk.

Archives of Disease in Childhood Fetal & Neonatal Edition 2008,93:F384–F388

38

Abstract

Background: Unconjugated hyperbilirubinemia occurs in almost all premature infants and is potentially neurotoxic. Treatment is based on total serum bilirubin (TSB), but treatment thresholds are not evidence based. Free bilirubin (Bf), i.e. not bound to albumin, seems a better parameter for bilirubin neurotoxicity, but measurements of Bf are not available in clinical practice. The bilirubin/albumin (B/A) ratio is con-sidered as a surrogate parameter for Bf and as an interesting additional parameter in the management of hyperbilirubinemia.

Objective: We reviewed the literature on the use of B/A ratios for predicting bilirubin-induced neurological dysfunction (BIND) including neurodevelopmental delay in jaundiced premature infants (gestational age less than 32 weeks).

Results: We performed a literature search and reviewed 6 publications regarding B/ A ratios in the management and outcome of jaundiced premature infants. No prospec-tive clinical trials exist demonstrating that bilirubin-induced neurotoxicity is reduced or that unnecessary treatment is avoided by using the B/A ratio in addition to TSB. Recently, a randomized controlled trial evaluating the effect of the additional use of the B/A ratio on neurodevelopmental outcome in jaundiced premature infants has been initiated.

Conclusion: Based on the prevailing evidence many authorities suggest that the ad-ditional use of the B/A ratio may be valuable when evaluating jaundiced premature infants. 

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 39

Background

Neonatal jaundice due to unconjugated hyperbilirubinemia occurs in almost all premature infants. Unconjugated hyperbilirubinemia is potentially harmful for the central nervous system and may result in kernicterus, causing severe and permanent neurologic sequelae.(3,4) Alternatively, more subtle encephalopathy, i.e. bilirubin-induced neurological dysfunction (BIND), including developmental delay may be due to hyperbilirubinemia.(5,6) Treatment of hyperbilirubinemia has been based on total serum bilirubin (TSB) concentration, but this has not been evidence based; explicit data on harmful ‘TSB-thresholds’ are lacking.(7) The risk of kernicterus and BIND may be in part determined by TSB, but also by the level of non-albumin bound free bilirubin (Bf), which can easily pass the blood-brain barrier, and thus better reflects the bilirubin load distributed in the brain.(8) Data from in vitro and in vivo experiments have suggested that Bf, and not TSB, is the principal determinant of bilirubin neurotoxicity.(9–11) Yet, multiple other factors (i.e., cerebral blood flow, vascular permeability, intactness of blood brain barrier (BBB), cellular efflux pumps) may affect the development of neurotoxicity at any given Bf level.(12)

Bilirubin-induced neurotoxicity may depend on the mutual relation between bilirubin and albumin as their levels and the intrinsic albumin-bilirubin binding constant (Ka) determine the Bf concentration, expressed in the equation:

TotalSerumBilirubin – FreeBilirubine

Albumine

K • FreeBilirubine

1 + (K • FreeBilirubine)=

The equation is helpful in illustrating the potential value of the TSB/ albumin (B/A) ratio considering the notion that TSB levels (expressed in µmol/L) are far higher than levels of Bf (expressed in µmol/L) so that the left half of the equation essentially is equal to the B/A ratio, which can easily be calculated, because TSB and albumin are measured using routine laboratory techniques. Many authorities advocate that the B/A ratio provides an additional assessment, i.e. next to but not instead of TSB, on which to base the management of hyperbilirubinemia.

Usefulness of the B/A ratio may be limited because many factors influence Ka; Ka may be decreased by drugs (e.g. ceftriaxon) or by plasma constituents (e.g. free fatty acids) that interfere with albumin-bilirubin binding.(13,14) Interindividual variations in Ka and in levels of albumin as well as binding of TSB to alternative binding sites render estimations of Bf predicted by Ka and B/A ratio at best a capricious exercise.

In premature infants, especially in very low birth weight (VLBW) infants (birth weight < 1500 g), bilirubin neurotoxicity may occur at lower TSB concentrations as

40

compared with term infants. We reviewed the theoretical considerations and clinical evidence for the concept that additional use of the B/A ratio in jaundiced prema-ture infants might improve the prediction of BIND including neurodevelopmental impairment.

Objectives

To determine the evidence supporting the use of B/A ratio in predicting BIND in-cluding neurodevelopmental impairment in premature infants (gestational age less than 32 weeks) with unconjugated hyperbilirubinemia.

Search strategy

Two authors independently searched the National Library of Medicine (Medline), Cochrane Library, including, Current Controlled Trials (CCT), ClinicalTrials.gov and NHS Centre for Reviews and Dissemination (CRD) from 1966 until July 2007. Search criteria included the following MESH and free text search terms: 1) infant, premature or low birth weight, 2) jaundice, neonatal, 3) hyperbilirubinemia or biliru-bin, 4) albumin, 5) bilirubin/ albumin ratio or B/A ratio, 6) (neurodevelopmental) outcome, 7) kernicterus, 8) bilirubin-induced neurological dysfunction or BIND, and 9) neurotoxicity.

Results

We found maximal 293 references following the search containing one of the com-binations from terms 1–3, 4 or 5, and 6–9 and screened the abstracts for relevance. The majority of the abstracts and papers that have been screened did not contain any outcome measure related to neurotoxicity or neurodevelopmental outcome; others did not contain albumin or B/A ratio measurements. All English-written papers with a combination of bilirubin or hyperbilirubinemia, albumin or B/A ratio, prematurity and an outcome related to neurotoxicity (including kernicterus) were included. Ultimately, 14 papers were eligible according to our selection criteria. Five (review) articles were found that provided opinion, but no data on the use of the B/A ratio in the management of hyperbilirubinemia in premature infants.(15–19) Three additional papers were excluded. One study concluded that the B/A ratio is a reliable surrogate for the Bf concentration and that the B/A ratio is a simple mode to incorporate the serum albumin concentration into exchange transfusion criteria, but provided no outcome data.(20) Another excluded study investigated the relation

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 41

between neonatal hyperbilirubinemia and several measures of psychoeducational outcome, including the Kaufman Mental Processing Composite (KMPC). Although albumin-determined binding capacity (calculated as 5(albumin)+2) was significantly and positively correlated with the KMPC, albumin and bilirubin levels were not explicitly provided. Thus, an inverse relationship between B/A ratio and outcome could only be speculated upon.(21) The third study was excluded, because part of the data were found to be similar to those of an already included publication.(22,23) Six articles were finally included. We found no prospective clinical trials evaluating the effic acy of B/A ratio, B/A ratio + TSB, or TSB in management or predicting outcome in premature infants with hyperbilirubinemia. One prospective cohort study evalu-ated short term effects of B/A ratios in predicting bilirubin encephalopathy,(24) and one study retrospectively examined the data of a prospective randomised controlled NICHD phototherapy trial to evaluate the association between neurodevelopmental outcome and the B/A ratio.(25–28) Two retrospective case-control post-mortem studies analyzed the relationship between the B/A ratio and kernicterus in prema-ture infants.(29–31) Risk factors for kernicterus in VLBW infants were described in a prospective study.(32) One article reported MRI documented kernicterus and the relation with B/A ratios.(33)

Amin et al. examined the usefulness of the B/A ratio and free bilirubin (Bf) as compared with TSB in predicting acute bilirubin encephalopathy in 143 infants of 28–32 weeks of gestational age.(34) Bilirubin encephalopathy was assessed by maturation of the auditory brainstem responses (ABR) in this single centre prospective cohort study. The mean peak TSB concentration in infants with normal ABR maturation was not significantly different from the mean peak TSB in infants with abnormal ABR maturation and there was a trend suggesting the B/A ratio (p=0.19) was better than TSB (p=0.98) in predicting abnormal ABR maturation. In a subset of 45 infants in whom Bf was measured, higher Bf (10.6 vs. 6.8 µmol/L, abnormal vs. normal ABR, resp.; p < 0.01) and higher B/A ratios (0.39 vs. 0.33 molar ratios, abnormal vs. normal ABR, resp; p < 0.05) preceded abnormal ABR maturation, whereas TSB did not. Bf proved to be the most sensitive predictor of transient bilirubin encephalopathy in premature newborns with hyperbilirubinemia. A Bf level of > 8.55 µmol/L had a relative risk of abnormal ABR of 2.45 (95% CI: 1.33–4.49). When Bf levels are not at hand, the authors suggest considering the B/A ratio along with the TSB in the management of hyperbilirubinemia in premature infants.

The National Institute of Child Health and Human Development Cooperative Phototherapy Study was performed in the US between 1974 and 1976.(35–38) This randomized controlled trial (n=1339) aimed to compare the efficacy of (prophylactic) phototherapy and/or exchange transfusion. The controls did not receive phototherapy

42

and had TSB levels maintained below specified levels using exchange transfusions. The study included a cohort of low birth weight (LBW) infants (< 2000 g; n=922). Phototherapy was considered safe and at least as effective as exchange transfusions only; neurodevelopmental outcome at 6 years of age was similar in both treatment groups.(39)

Scheidt et al. evaluated neurodevelopmental outcome and TSB levels in a sub-group of 224 premature / LBW infants who were randomized to the control group (no phototherapy, TSB below specified levels using exchange transfusions).(40) Neither cerebral palsy, nor IQ was significantly associated with TSB levels, time, or duration of exposure to bilirubin. IQ decreased at higher B/A ratios (r=-0.12; p=0.06) but the relationship disappeared after correction for neonatal risk factors.

Kim et al. reviewed 398 neonatal autopsies in infants.(41) Kernicterus, defined as yellowish staining of cerebral gray matter, was reported in 27 of 398 (7%) autopsied infants. Clinical and pathologic factors in these 27 kernicteric infants (32 weeks and 1417 g, mean GA and BW, respectively.) were retrospectively compared to a control group of 103 autopsied infants with similar birth weight and gestational age, but without kernicterus. Kernicteric infants had rather low mean TSB peak concentra-tions, which were similar to those in the infants without kernicterus (197 ± 55 vs. 205 ± 96 µmol/L). Serum albumin values and the reserve albumin binding capacity were significantly lower in kernicteric infants compared to controls. B/A ratios could be calculated in 6 of the 27 kernicteric infants and in 15 of the 103 non-kernicteric infants, and appeared similar (0.44 ± 0.21 vs. 0.44 ± 0.36, kernicteric vs. non-kernicteric in-fants, respectively).

Cashore et al. determined TSB, Bf, B/A ratio and bilirubin binding affinity in 13 pre-mature infants (< 1500 gram) with hyperbilirubinemia before exchange transfusions were performed and who died in the neonatal period.(42) Five of these 13 infants had kernicterus, defined as yellowish staining of the basal ganglia or brainstem at autopsy. Compared to non-kernicteric infants, Bf levels were increased in infants with kernic-terus (27 ± 9 vs. 13 ± 10 µmol/L, kernicteric vs. non-kernicteric infants respectively; p< 0.05). Moreover, bilirubin binding capacities expressed as moles of bilirubin per mole of albumin, i.e. molar B/A ratios, were lower in infants with kernicterus (0.42 ± 0.07 vs. 0.54 ± 0.08, kernicteric vs. non-kernicteric infants, respectively; p< 0.05).

Ritter et al. prospectively assessed risk factors in the development of kernicterus in 91 VLBW infants.(43) Peak TSB, and Bf or albumin did not significantly differ between infants with and without kernicterus. Although not explicitly provided in their paper, B/A ratios appeared higher in VLBW infants with kernicterus (calculated group mean values 2.6 vs. 2.2 mg/g, kernicteric vs. non-kernicteric VLBW infants, respectively). Methodological shortcomings, e.g. measuring Bf in 40-fold diluted

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 43

Tabl

e 1. P

ublic

atio

ns re

gard

ing t

he u

se o

f B/A

ratio

s and

out

com

e in

prem

atur

e inf

ants

with

unc

onju

gate

d hy

perb

iliru

bine

mia

.

Auth

or/

year

Stu

dy

desi

gn

Stu

dy

pop

ulati

on

Met

hod

Rela

tion B

/A r

ati

o*

and

outc

om

e

Am

in,

200

1(89

)Pr

osp

ect

ive c

ohort

st

ud

y14

3 in

fants

(G

A: 2

8–3

2

weeks

)M

atu

rati

on o

f th

e a

ud

itory

b

rain

stem

resp

onse

s (A

BR)

Hig

her

B/A

rati

os

pre

ced

ed

ab

norm

al A

BR m

atu

rati

on

Sch

eid

t,

199

1(9

0)

Ret

rosp

ect

ive a

na

ly-

sis

of a

RC

T22

4 p

rem

atu

re /

LBW

in

fants

(<20

00

g)

Neuro

log

ica

l exa

min

ati

on,

Wech

sler

Inte

llig

ence

Sca

le

for

Child

ren-R

evis

ed

at

6 y

ears

IQ d

ecr

ease

d a

t hig

her

B/A

ra

tios,

no r

ela

tionsh

ip a

fter

co

rrect

ion for

neonata

l ris

k fa

ctors

Kim

, 19

80

(91)

Case

-contr

ol p

ost

-m

ort

em

stu

dy

27 k

ernic

teri

c in

fants

(G

A:

27–3

8 w

eeks

) and

10

3

matc

hed

contr

ols

Com

pari

son o

f cl

inic

al,

lab

ora

-to

ry a

nd h

isto

path

olo

gic

al

data

Low

er a

lbum

in le

vels

in k

er-

nic

teri

c in

fants

, but

ava

ilab

le

B/A

rati

os

(6 o

f 27

and

15 o

f 10

3)

ap

peare

d s

imila

r

Cash

ore

, 19

82(

92)

Case

-contr

ol p

ost

-m

ort

em

stu

dy

13 p

rem

atu

re in

fants

(<

150

0 g

ram

); 5

wit

h

kern

icte

rus

Com

pari

son o

f la

bora

tory

d

ata

Bind

ing

cap

aci

ties

exp

ress

ed

as

mola

r B/

A r

ati

os

wer

e lo

wer

in

infa

nts

wit

h k

ernic

teru

s

Ritt

er,

198

2(9

3)

Prosp

ect

ive c

ohort

st

ud

y9

1 p

rem

atu

re V

LBW

infa

nts

(<

150

0g

)A

ssess

ment

of r

isk

fact

ors

in

the d

evelo

pm

ent

of k

ernic

-te

rus

B/A

rati

os

ap

peare

d h

igher

in

infa

nts

wit

h k

ernic

teru

s#

Gova

ert,

20

03(9

4)

Ret

rosp

ect

ive c

ase

st

ud

y5 p

rem

atu

re (

GA

: 25 t

o 2

9

weeks

) and

3 ter

m in

fants

w

ith c

linic

al s

igns

of k

ernic

-te

rus

ABR

, ma

gnet

ic r

eso

nance

im-

ag

ing

(M

RI)

and

/or

ultr

aso

und

Ab

norm

al A

BR, M

RI a

nd/o

r ul

traso

und

in a

ll p

rem

atu

re in

-fa

nts

wit

h e

leva

ted

B/A

rati

os

*: B

/A r

ati

os

wer

e c

alc

ulate

d f

rom

ser

um

bili

rub

in a

nd a

lbum

in c

once

ntr

ati

ons,

whic

h w

ere m

easu

red

by

routi

ne la

bora

tory

tech

niq

ues.

In t

he s

tud

y

of C

ash

ore

, bili

rub

in b

ind

ing

cap

aci

ties

wer

e e

stim

ate

d f

rom

bili

rub

in t

itrati

on c

urv

es

and

exp

ress

ed

as

mole

s of

bili

rub

in p

er m

ole

of

alb

um

in, i

.e.

mola

r B/

A r

ati

os

at

ap

pare

nt

satu

rati

on o

f firs

t b

ind

ing

site o

f a

lbum

in. A

ccord

ing

ly, l

ow B

/A r

ati

os

reflect

low

bili

rub

in b

ind

ing

cap

aci

ties

in t

he la

t-te

r st

ud

y.

#: d

ata

not

pro

vid

ed

, but

calc

ulate

d b

y p

rese

nti

ng

auth

ors

44

samples according to the original peroxidase method of Jacobsen and not correcting for rate limiting dissociation of bilirubin from albumin, may have prevented the dif-ferences in Bf levels (but not the TSB) from reaching statistically significance.(44,45)

Govaert et al. reported five premature infants (25 to 29 weeks gestational age) with clinical signs of kernicterus and long term changes in globus pallidus on magnetic resonance imaging (MRI) and/or sonography.(46) Although TSB levels remained below exchange transfusion thresholds, B/A ratios were high in these infants. ABR were severely impaired in all premature infants with elevated B/A ratios and all developed hearing loss. In three infants, a combined respiratory and metabolic aci-dosis had been present around the peak TSB level. The authors conclude that the pathophysiological role of low serum albumin levels must be considered in BIND, especially when acidosis and jaundice are present in premature infants.

Discussion

Premature infants are more prone to neurological impairment as well as bilirubin neurotoxicity than their term counterparts.(47–50) Recently, the role of albumin and bilirubin-albumin binding in the pathophysiology of bilirubin-induced neurotoxicity in premature infants has been discussed.(51–54) Several advisory committees and experts in the field of bilirubin research advocate the additional use of B/A ratios in jaundiced premature (or term infants), especially when the TSB level is close to that at which exchange transfusion is recommended.(55–62) We found no published prospective clinical trials documenting that the use of B/A ratios (or TSB for that mat-ter) in the management of hyperbilirubinemia reduces long term bilirubin-induced neurotoxicity. The relationship between TSB and neurodevelopmental outcome is unclear and has resulted in guidelines with arbitrary TSB thresholds for interven-tion in premature as well as in term infants with unconjugated hyperbilirubinemia.(63–66) Compared with TSB, Bf has been found a more reliable predictor of bilirubin neurotoxicity as assessed by ABR maturation and electroencephalography and to cor-relate better with long term outcome (kernicterus).(67–70) Unfortunately, routine clinical laboratory measurements of Bf are not generally available. B/A ratios can be calculated based on standard laboratory assays and the B/A ratio has been more useful than TSB alone to indicate the Bf concentration.(71,72)

However, due to the presence of drugs that interfere with bilirubin binding to albumin, Bf may be much higher than suggested by the calculated B/A ratio and the intrinsic affinity of the albumin for bilirubin. Furthermore, other plasma constituents such as apolipoproteins bind unconjugated bilirubin. Consequently, the B/A ratio seems an imperfect surrogate to estimate Bf, but even Bf is likely to

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 45

be an imperfect predictor of neurotoxicity; many (downstream) factors affect the development of neurotoxicity at any given Bf level. Bf in plasma is always available for diffusion across the blood-brain barriers, at rates that depend on the gradient compared to the unmeasured Bf levels in the brain and cerebrospinal fluid (CSF). The B/A ratio need not be very high for Bf in plasma to exceed Bf in the brain or CSF. Bf in cerebral tissue ultimately determines the development of UCB neurotoxicity. Protective mechanisms moderating neurotoxicity include: 1) limited binding to GSH (reduced glutathione)-transferases in the brain or albumin in the CSF which impairs ongoing movement of bilirubin to the brain, 2) membrane-bound transport-ers (e.g., the multi-drug resistance protein 1 (MDR1), multi-drug resistance related protein 1 (MRP1)) which are involved in bilirubin efflux from the central nervous system, and 3) intracellular metabolism of UCB, i.e. conjugation and/or oxidation of UCB.(73–75) Development of toxicity may depend also on protective anti-oxidant and anti-apoptotic mechanisms, which may differ among cell types. In vitro and in vivo studies demonstrating the neurotoxic potential of Bf are often confounded by solubility issues and poor delineation of the Bf present.(76–78) Contrary to earlier reports,(6,30) in vitro toxicity to neurons and astrocytes may occur well below aque-ous saturation of Bf (70 nM) if protective mechanisms, e.g. activity of MRP1, are impaired.(79,80) Self-aggregation of UCB is, therefore, not a necessary condition for UCB toxicity. These studies have recently been reviewed;(81–83) although Bf is the measurable principal determinant of bilirubin neurotoxicity, more information is needed on the approximate levels of Bf associated with neurotoxicity.

Experts in neonatal jaundice have suggested that – in the absence of commercial assays for Bf – the B/A ratio can be used in conjunction with TSB in the evaluation and treatment of premature infants with hyperbilirubinemia.(84–88) It is currently mainly theoretical considerations that favor the use of the B/A ratio in addition to TSB levels, considering the rather small number of publications regarding B/A ratio and outcome and the lack of prospective clinical trials supporting the clinical benefit, i.e. prevention or reduction of BIND in premature infants. The use of B/A ratios (and/or Bf) in addition to TSB measurements needs to be evaluated in premature newborns in randomized clinical trials which include assessment of neurodevelopmental out-come and/ or ABR and MRI for detecting and documenting bilirubin toxicity. In line with this, a prospective randomized controlled trial has started in April 2007 in the Netherlands to evaluate the additional use of the B/A ratio in jaundiced premature infants. Primary outcome of this so-called Bilirubin Albumin Ratio Trial (BARTrial) is neurodevelopmental outcome at 18–24 months corrected age (http://www.trial-register.nl/trialreg/admin/rctview.asp?TC=935). Treatment thresholds of TSB and

46

B/A ratio in standard and high risk infants for intervention with phototherapy and exchange transfusion in this continuing RCT are shown in Table 2.

Table 2. Treatment thresholds* of TSB (µmol/L) and B/A ratio (µmol/g) for intervention with phototherapy and exchange transfusion in standard and high risk premature infants included in the Bilirubin Albumin Ratio Trial (BARTrial).

Birthweight (g) Phototherapy Exchange transfusion

Standard risk High risk Standard risk High risk

TSB B/A TSB B/A TSB B/A TSB B/A

<1250 150 6.0 100 4.0 220 8.8 170 6.8

1250–1499 190 6.3 150 5.0 260 10.4 220 8.8

1500–2000 220 7.3 190 6.3 290 11.6 260 10.4

2000–2500 240 8.0 220 7.3 310 12.4 290 11.6

*: adapted from Ahlfors (95) and Maisels (96)

Inclusion criteria: gestational age < 32 weeks without chromosomal or syndromal abnormalities.

Intervention: Hyperbilirubinemia is evaluated daily and treatment is based on B/A ratio and TSB (whichever comes first) in the study group versus TSB only in the control group.

Risk factors: asphyxia, hypoxemia, acidosis, haemolysis, extreme low birth weight, sepsis/ meningitis and intracranial haemorrhage.

Conclusion

The need for additional parameters for the use of phototherapy and exchange transfu-sion in jaundiced premature infants is generally agreed upon; the wide range of cur-rently used ‘TSB-thresholds’ are arbitrary and poorly evidenced based as is reflected by the poor relationship between TSB levels and neurodevelopmental outcome. Awaiting prospective data of a recently started RCT, available data suggests that the concurrent use of B/A ratio and TSB in the management of hyperbilirubinemia may provide a tool for the development of more robust criteria for managing newborn jaundice and is likely to be better than TSB alone.

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 47

References

1. Morris BH, Oh W, Tyson JE, Stevenson DK, Phelps DL, O’Shea TM, et al. Aggressive vs. conservative phototherapy for infants with extremely low birth weight. N Engl J Med 2008 Oct 30,359(18):1885–96.

2. Jangaard KA, Vincer MJ, Allen AC. A randomized trial of aggressive versus conservative phototherapy for hyperbilirubinemia in infants weighing less than 1500 g: Short- and long-term outcomes. Paediatr Child Health 2007 Dec;12(10):853–8.

3. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F455-F458.

4. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004 Jul;114(1):297–316.

5. Oh W, Tyson JE, Fanaroff AA, Vohr BR, Perritt R, Stoll BJ, et al. Association Between Peak Serum Bilirubin and Neurodevelopmental Outcomes in Extremely Low Birth Weight Infants. Pediatrics 2003 Oct 1,112(4):773–9.

6. Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 2005 Jan;25(1):54–9.

7. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

8. Ostrow JD, Pascolo L, Shapiro SM, Tiribelli C. New concepts in bilirubin encephalopathy. Eur J Clin Invest 2003 Nov;33(11):988–97.

9. Ahlfors CE, Shapiro SM. Auditory brainstem response and unbound bilirubin in jaundiced (jj) Gunn rat pups. Biol Neonate 2001 Aug;80(2):158–62.

10. Ostrow JD, Pascolo L, Tiribelli C. Reassessment of the unbound concentrations of unconjugated bilirubin in relation to neurotoxicity in vitro. Pediatr Res 2003 Jul;54(1):98–104.

11. Calligaris SD, Bellarosa C, Giraudi P, Wennberg RP, Ostrow JD, Tiribelli C. Cytotoxicity is pre-dicted by unbound and not total bilirubin concentration. Pediatr Res 2007 Nov;62(5):576–80.

12. Wennberg R, Ahlfors C, Bhutani V, Johnson L, Shapiro S. Toward Understanding Kernicterus: A Challenge to Improve the Management of Jaundiced Newborns. Pediatrics 2006 Feb 1,117(2):474–85.

13. Odell GB, Cukier JO, Ostrea EM, Jr., Maglalang AC, Poland RL. The influence of fatty acids on the binding of bilirubin to albumin. J Lab Clin Med 1977 Feb;89(2):295–307.

14. Robertson A, Karp W, Brodersen R. Bilirubin displacing effect of drugs used in neonatology. Acta Paediatr Scand 1991 Dec;80(12):1119–27.

15. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

16. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F455-F458.

17. Bhutani VK, Johnson LH. Urgent clinical need for accurate and precise bilirubin measurements in the United States to prevent kernicterus. Clin Chem 2004 Mar;50(3):477–80.

48

18. Kaplan M, Hammerman C. Understanding severe hyperbilirubinemia and preventing kernicterus: adjuncts in the interpretation of neonatal serum bilirubin. Clin Chim Acta 2005 Jun;356(1–2):9–21.

19. Bhutani VK, Johnson LH, Shapiro SM. Kernicterus in sick and preterm infants (1999–2002): a need for an effective preventive approach. Semin Perinatol 2004 Oct;28(5):319–25.

20. Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns. Pediatrics 1994 Mar;93(3):488–94.

21. Hansen RL, Hughes GG, Ahlfors CE. Neonatal bilirubin exposure and psychoeducational outcome. J Dev Behav Pediatr 1991 Oct;12(5):287–93.

22. Cashore WJ. Free bilirubin concentrations and bilirubin-binding affinity in term and preterm infants. J Pediatr 1980 Mar;96(3 Pt 2):521–7.

23. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low-birth-weight infants. Pediatrics 1982 Apr;69(4):481–5.

24. Amin SB, Ahlfors C, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 2001 Apr;107(4):664–70.

25. Brown AK, Kim MH, Wu PY, Bryla DA. Efficacy of phototherapy in prevention and management of neonatal hyperbilirubinemia. Pediatrics 1985 Feb;75(2 Pt 2):393–400.

26. Bryla DA. Randomized, controlled trial of phototherapy for neonatal hyperbilirubinemia. Development, design, and sample composition. Pediatrics 1985 Feb;75(2 Pt 2):387–92.

27. Scheidt PC, Bryla DA, Nelson KB, Hirtz DG, Hoffman HJ. Phototherapy for neonatal hyperbilirubinemia: six-year follow-up of the National Institute of Child Health and Human Development clinical trial. Pediatrics 1990 Apr;85(4):455–63.

28. Scheidt PC, Graubard BI, Nelson KB, Hirtz DG, Hoffman HJ, Gartner LM, et al. Intelligence at six years in relation to neonatal bilirubin levels: follow-up of the National Institute of Child Health and Human Development Clinical Trial of Phototherapy. Pediatrics 1991 Jun;87(6):797–805.

29. Kim MH, Yoon JJ, Sher J, Brown AK. Lack of predictive indices in kernicterus: a comparison of clinical and pathologic factors in infants with or without kernicterus. Pediatrics 1980 Dec;66(6):852–8.

30. Cashore WJ. Free bilirubin concentrations and bilirubin-binding affinity in term and preterm infants. J Pediatr 1980 Mar;96(3 Pt 2):521–7.

31. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low-birth-weight infants. Pediatrics 1982 Apr;69(4):481–5.

32. Ritter DA, Kenny JD, Norton HJ, Rudolph AJ. A prospective study of free bilirubin and other risk factors in the development of kernicterus in premature infants. Pediatrics 1982 Mar;69(3):260–6.

33. Govaert P, Lequin M, Swarte R, Robben S, De Coo R, Weisglas-Kuperus N, et al. Changes in globus pallidus with (pre)term kernicterus. Pediatrics 2003 Dec;112(6 Pt 1):1256–63.

34. Amin SB, Ahlfors C, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 2001 Apr;107(4):664–70.

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 49

35. Brown AK, Kim MH, Wu PY, Bryla DA. Efficacy of phototherapy in prevention and management of neonatal hyperbilirubinemia. Pediatrics 1985 Feb;75(2 Pt 2):393–400.

36. Bryla DA. Randomized, controlled trial of phototherapy for neonatal hyperbilirubinemia. Development, design, and sample composition. Pediatrics 1985 Feb;75(2 Pt 2):387–92.

37. Scheidt PC, Bryla DA, Nelson KB, Hirtz DG, Hoffman HJ. Phototherapy for neonatal hyperbilirubinemia: six-year follow-up of the National Institute of Child Health and Human Development clinical trial. Pediatrics 1990 Apr;85(4):455–63.

38. Scheidt PC, Graubard BI, Nelson KB, Hirtz DG, Hoffman HJ, Gartner LM, et al. Intelligence at six years in relation to neonatal bilirubin levels: follow-up of the National Institute of Child Health and Human Development Clinical Trial of Phototherapy. Pediatrics 1991 Jun;87(6):797–805.

39. Scheidt PC, Bryla DA, Nelson KB, Hirtz DG, Hoffman HJ. Phototherapy for neonatal hyperbilirubinemia: six-year follow-up of the National Institute of Child Health and Human Development clinical trial. Pediatrics 1990 Apr;85(4):455–63.

40. Scheidt PC, Graubard BI, Nelson KB, Hirtz DG, Hoffman HJ, Gartner LM, et al. Intelligence at six years in relation to neonatal bilirubin levels: follow-up of the National Institute of Child Health and Human Development Clinical Trial of Phototherapy. Pediatrics 1991 Jun;87(6):797–805.

41. Kim MH, Yoon JJ, Sher J, Brown AK. Lack of predictive indices in kernicterus: a comparison of clinical and pathologic factors in infants with or without kernicterus. Pediatrics 1980 Dec;66(6):852–8.

42. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low-birth-weight infants. Pediatrics 1982 Apr;69(4):481–5.

43. Ritter DA, Kenny JD, Norton HJ, Rudolph AJ. A prospective study of free bilirubin and other risk factors in the development of kernicterus in premature infants. Pediatrics 1982 Mar;69(3):260–6.

44. Jacobsen J, Wennberg RP. Determination of unbound bilirubin in the serum of newborns. Clin Chem 1974 Jul;20(7):783.

45. Ahlfors CE. Benzyl alcohol, kernicterus, and unbound bilirubin. J Pediatr 2001 Aug;139(2):317–9.

46. Govaert P, Lequin M, Swarte R, Robben S, De Coo R, Weisglas-Kuperus N, et al. Changes in globus pallidus with (pre)term kernicterus. Pediatrics 2003 Dec;112(6 Pt 1):1256–63.

47. Gartner LM, Snyder RN, Chabon RS, Bernstein J. Kernicterus: high incidence in premature infants with low serum bilirubin concentrations. Pediatrics 1970 Jun;45(6):906–17.

48. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low-birth-weight infants. Pediatrics 1982 Apr;69(4):481–5.

49. Watchko JF, Oski FA. Kernicterus in preterm newborns: past, present, and future. Pediatrics 1992 Nov;90(5):707–15.

50. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F455-F458.

50

51. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

52. Ahlfors CE, Parker AE. Evaluation of a model for brain bilirubin uptake in jaundiced newborns. Pediatr Res 2005 Dec;58(6):1175–9.

53. Kaplan M, Hammerman C. Understanding severe hyperbilirubinemia and preventing kernicterus: adjuncts in the interpretation of neonatal serum bilirubin. Clin Chim Acta 2005 Jun;356(1–2):9–21.

54. Wennberg R, Ahlfors C, Bhutani V, Johnson L, Shapiro S. Toward Understanding Kernicterus: A Challenge to Improve the Management of Jaundiced Newborns. Pediatrics 2006 Feb 1,117(2):474–85.

55. Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns. Pediatrics 1994 Mar;93(3):488–94.

56. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F455-F458.

57. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

58. Blackmon LR, Fanaroff AA, Raju TN. Research on prevention of bilirubin-induced brain injury and kernicterus: National Institute of Child Health and Human Development conference executive summary. 2003. Pediatrics 2004 Jul;114(1):229–33.

59. Bhutani VK, Johnson LH. Urgent clinical need for accurate and precise bilirubin measurements in the United States to prevent kernicterus. Clin Chem 2004 Mar;50(3):477–80.

60. Bhutani VK, Johnson LH, Shapiro SM. Kernicterus in sick and preterm infants (1999–2002): a need for an effective preventive approach. Semin Perinatol 2004 Oct;28(5):319–25.

61. Kaplan M, Hammerman C. Understanding severe hyperbilirubinemia and preventing kernicterus: adjuncts in the interpretation of neonatal serum bilirubin. Clin Chim Acta 2005 Jun;356(1–2):9–21.

62. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004 Jul;114(1):297–316.

63. Oh W, Tyson JE, Fanaroff AA, Vohr BR, Perritt R, Stoll BJ, et al. Association Between Peak Serum Bilirubin and Neurodevelopmental Outcomes in Extremely Low Birth Weight Infants. Pediatrics 2003 Oct 1,112(4):773–9.

64. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F455-F458.

65. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

66. Wennberg R, Ahlfors C, Bhutani V, Johnson L, Shapiro S. Toward Understanding Kernicterus: A Challenge to Improve the Management of Jaundiced Newborns. Pediatrics 2006 Feb 1,117(2):474–85.

67. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low-birth-weight infants. Pediatrics 1982 Apr;69(4):481–5.

Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity 51

68. Nakamura H, Yonetani M, Uetani Y, Funato M, Lee Y. Determination of serum unbound bilirubin for prediction of kernicterus in low birthweight infants. Acta Paediatr Jpn 1992 Dec;34(6):642–7.

69. Funato M, Tamai H, Shimada S, Nakamura H. Vigintiphobia, unbound bilirubin, and auditory brainstem responses. Pediatrics 1994 Jan;93(1):50–3.

70. Amin SB, Ahlfors C, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 2001 Apr;107(4):664–70.

71. Ahlfors CE, Wennberg RP. Bilirubin-albumin binding and neonatal jaundice. Semin Perinatol 2004 Oct;28(5):334–9.

72. Ahlfors CE, Parker AE. Evaluation of a model for brain bilirubin uptake in jaundiced newborns. Pediatr Res 2005 Dec;58(6):1175–9.

73. Hansen TW. Bilirubin oxidation in brain. Mol Genet Metab 2000 Sep;71(1–2):411–7.

74. Tiribelli C, Ostrow JD. The molecular basis of bilirubin encephalopathy and toxicity: Report of an EASL Single Topic Conference, Trieste, Italy, 1–2 October, 2004. Journal of Hepatology 2005 Jul;43(1):156–66.

75. Wennberg R, Ahlfors C, Bhutani V, Johnson L, Shapiro S. Toward Understanding Kernicterus: A Challenge to Improve the Management of Jaundiced Newborns. Pediatrics 2006 Feb 1,117(2):474–85.

76. Ostrow JD, Pascolo L, Tiribelli C. Reassessment of the unbound concentrations of unconjugated bilirubin in relation to neurotoxicity in vitro. Pediatr Res 2003 Jul;54(1):98–104.

77. Ostrow JD, Pascolo L, Shapiro SM, Tiribelli C. New concepts in bilirubin encephalopathy. Eur J Clin Invest 2003 Nov;33(11):988–97.

78. Ostrow JD, Pascolo L, Brites D, Tiribelli C. Molecular basis of bilirubin-induced neurotoxicity. Trends Mol Med 2004 Feb;10(2):65–70.

79. Gennuso F, Fernetti C, Tirolo C, Testa N, L’Episcopo F, Caniglia S, et al. Bilirubin protects astrocytes from its own toxicity by inducing up-regulation and translocation of multidrug resistance-associated protein 1 (Mrp1). Proc Natl Acad Sci U S A 2004 Feb 24,101(8):2470–5.

80. Falcao AS, Bellarosa C, Fernandes A, Brito MA, Silva RFM, Tiribelli C, et al. Role of multidrug resistance-associated protein 1 expression in the in vitro susceptibility of rat nerve cell to unconjugated bilirubin. Neuroscience 2007 Feb 9,144(3):878–88.

81. Ostrow JD, Pascolo L, Shapiro SM, Tiribelli C. New concepts in bilirubin encephalopathy. Eur J Clin Invest 2003 Nov;33(11):988–97.

82. Ostrow JD, Pascolo L, Tiribelli C. Reassessment of the unbound concentrations of unconjugated bilirubin in relation to neurotoxicity in vitro. Pediatr Res 2003 Jul;54(1):98–104.

83. Tiribelli C, Ostrow JD. The molecular basis of bilirubin encephalopathy and toxicity: Report of an EASL Single Topic Conference, Trieste, Italy, 1–2 October, 2004. Journal of Hepatology 2005 Jul;43(1):156–66.

84. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

52

85. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F455-F458.

86. Bhutani VK, Johnson LH, Shapiro SM. Kernicterus in sick and preterm infants (1999–2002): a need for an effective preventive approach. Semin Perinatol 2004 Oct;28(5):319–25.

87. Bhutani VK, Johnson LH. Urgent clinical need for accurate and precise bilirubin measurements in the United States to prevent kernicterus. Clin Chem 2004 Mar;50(3):477–80.

88. Kaplan M, Hammerman C. Understanding severe hyperbilirubinemia and preventing kernicterus: adjuncts in the interpretation of neonatal serum bilirubin. Clin Chim Acta 2005 Jun;356(1–2):9–21.

89. Amin SB, Ahlfors C, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 2001 Apr;107(4):664–70.

90. Scheidt PC, Graubard BI, Nelson KB, Hirtz DG, Hoffman HJ, Gartner LM, et al. Intelligence at six years in relation to neonatal bilirubin levels: follow-up of the National Institute of Child Health and Human Development Clinical Trial of Phototherapy. Pediatrics 1991 Jun;87(6):797–805.

91. Kim MH, Yoon JJ, Sher J, Brown AK. Lack of predictive indices in kernicterus: a comparison of clinical and pathologic factors in infants with or without kernicterus. Pediatrics 1980 Dec;66(6):852–8.

92. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low-birth-weight infants. Pediatrics 1982 Apr;69(4):481–5.

93. Ritter DA, Kenny JD, Norton HJ, Rudolph AJ. A prospective study of free bilirubin and other risk factors in the development of kernicterus in premature infants. Pediatrics 1982 Mar;69(3):260–6.

94. Govaert P, Lequin M, Swarte R, Robben S, De Coo R, Weisglas-Kuperus N, et al. Changes in globus pallidus with (pre)term kernicterus. Pediatrics 2003 Dec;112(6 Pt 1):1256–63.

95. Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns. Pediatrics 1994 Mar;93(3):488–94.

96. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 Nov;88(6):F459-F463.

97. Morris BH, Oh W, Tyson JE, Stevenson DK, Phelps DL, O’Shea TM, et al. Aggressive vs. conservative phototherapy for infants with extremely low birth weight. N Engl J Med 2008 Oct 30,359(18):1885–96.

98. Jangaard KA, Vincer MJ, Allen AC. A randomized trial of aggressive versus conservative phototherapy for hyperbilirubinemia in infants weighing less than 1500 g: Short- and long-term outcomes. Paediatr Child Health 2007 Dec;12(10):853–8.

99. Curtis-Cohen M, Stahl GE, Costarino AT, Polin RA. Randomized trial of prophylactic phototherapy in the infant with very low birth weight. J Pediatr 1985 Jul;107(1):121–4.

567

9SD

234

8

53

Chapter 4

A double-blind, randomized controlled

trial on the bilirubin/ albumin ratio in

jaundiced preterm infants

The BARTrial Study group

Submitted

54

Abstract

Background: High bilirubin/albumin (B/A) ratios increase the risk of bilirubin neu-rotoxicity. The B/A ratio may be a more appropriate measure than the total serum bilirubin (TSB) level to manage treatment of preterm infants with hyperbilirubinemia We aimed to assess whether the additional use of B/A ratio thresholds in manag-ing hyperbilirubinemia in preterms improved neurodevelopmental outcome.

Methods: In a prospective, randomized controlled trial, 615 preterms of 32 weeks’ gestation or less were randomly assigned to treatment based on either B/A ratio and TSB thresholds, whichever threshold was crossed first, or on TSB thresholds only. The primary outcome was neurodevelopment at 18 to 24 months’ corrected age as assessed with the Bayley Scales of Infant Development III by investigators unaware of treatment allocation. Secondary outcomes included complications of preterm birth and death.

Results: Composite motor and cognitive scores did not differ between the B/A ratio and TSB  groups. Demographic characteristics, maximal TSB levels, B/A ratios, and other secondary outcomes were similar. The rates of death and/or severe neurode-velopmental impairment for the two groups were 15.4% versus 15.5% (P=1.0) and 2.8% versus 1.4% (P=0.62) for birth weights ≤1000 g and 1.8% versus 5.8% (P=0.03) and 4.1% versus 2.0% (P=0.26) for birth weights of >1000 g.

Conclusions: The additional use of B/A ratios in hyperbilirubinemia management of preterms did not improve their neurodevelopmental outcome. More research is needed to confirm the reduction in mortality observed in infants with birth weights of >1000 g allocated to the B/A ratio group. (Clinical Trials: ISRCTN74465643).

The BARTrial 55

Background

Free bilirubin (Bf), i.e. the fraction of bilirubin not bound to albumin, crosses the blood-brain barrier and exhibits neurotoxicity. In accordance, Bf predicts biliru-bin neurotoxicity more reliably than total serum bilirubin (TSB) as assessed by clinical and electrophysiological parameters, i.e. neurodevelopmental outcome and maturation of automated brain stem responses, respectively.(1,2) Nevertheless, Bf is not incorporated in the clinical management of hyperbilirubinemia. Neither is it routine practice to measure Bf in clinical laboratories nor are FDA approved instru-ments available commercially. The bilirubin (TSB)/albumin (B/A) ratio, which is considered a surrogate parameter for Bf, might be an ideal candidate to accurately indicate an increased risk of bilirubin-induced neurotoxicity in preterm infants. To date, mainly retrospective data have favored an additional role for high B/A ratios as risk factors for bilirubin-induced neurotoxicity and only limited data exist regarding B/A ratios in the management and neurodevelopmental outcome of preterm infants with unconjugated hyperbilirubinemia.(3) What is lacking are prospective clinical trials on the use of the B/A ratio in jaundiced preterms to support its clinical benefit of preventing or reducing long-term bilirubin-induced neurotoxicity.

We conducted a prospective, multicenter, randomized controlled trial on the management of preterms with hyperbilirubinemia to determine the effects of us-ing B/A ratios on their neurodevelopmental outcome at 18 to 24 months after the expected date of delivery.

Methods

Patients

Preterm infants of 31+6 weeks’ gestation or less (based on the first day of the last menstruation), were eligible for enrollment after admission to a neonatal intensive care unit (NICU) during the first 24 hours after birth. Exclusion criteria were major congenital malformations, clinical syndromes, or chromosomal abnormalities. The Medical Ethics Review Board of the University Medical Center Groningen, the Netherlands and all ten Dutch NICUs approved the study. Written informed consent was obtained from the parents or guardian of each participating infant.

Enrollment and treatment

Infants were stratified on the basis of center and gestational age (24+0 to 28+6 and 29+0 to 31+6 weeks), and were randomly assigned to a treatment group (balance [1:1]) by a centralized computer system. During the first ten days hyperbilirubinemia in

56

infants in the B/A ratio group was evaluated daily using the B/A ratio together with TSB. Treatment decisions were based on the B/A ratio and the TSB level, whichever threshold was crossed first. The TSB and albumin levels were also evaluated daily for infants in the TSB group (control group), but treatment was based on the TSB level only. TSB and albumin were measured using routine analytical methods.(4)

The TSB treatment thresholds (in µmol/L) for phototherapy (PT) and exchange transfusion were uniform and consensus-based. The B/A ratio thresholds (in µmol/L divided by g/L = µmol/g) were also uniform and derived from the corresponding TSB values and albumin levels of 25 g/L and 30 g/L for infants with birth weights of < or >1250 g, respectively.(5) Treatment nomograms depended on infants’ birth weights and risk factors.(6)

Assessments

The primary outcome was the composite motor score at 18 to 24 months’ corrected age because previous retrospective data analyses of the National Institute of Child Health and Human Development (NICHD) Phototherapy Trial showed that peak bilirubin levels correlated with psychomotor developmental indices (PDI).(3) The 3rd version of the Bayley Scales of Infant and Toddler Development (BSID III) was used to assess fine and gross motor as well as cognitive functions by investigators who were unaware of the treatment allocations. Standardized, pediatric, neurologic evaluations were performed at the same age as the BSID III assessments. Cerebral palsy and mild neurological dysfunction were diagnosed using international criteria and definitions.(7) Hearing and vision were assessed with standardized tests.

Secondary outcomes included death, neurodevelopmental impairment (NDI), pre-defined clinical diagnoses, and potential adverse events. Neurodevelopmental impairment was defined as a composite motor score of <85 or a composite mental score of <85, any degree of cerebral palsy, any visual impairment, or any hearing impairment. Severe NDI was defined as a composite motor score of <70 or a com-posite mental score of <70, moderate or severe cerebral palsy, bilateral blindness, or bilateral hearing loss.

Statistical analysis

We considered a difference between the two treatment groups on the composite motor score ≥7 points clinically relevant. To detect this on a 100 point scale, SD 15, alpha 0.05, power 0.80, we planned to evaluate 434 infants. Accounting for 10% mor-tality and 20% incomplete follow-up, a total of 614 subjects needed to be included, i.e. 307 per group. Intentions to treat analyses were performed.

The BARTrial 57

We pre-specified two subgroups: birth weights ≤1000 g and birth weights of >1000 g with high B/A ratios. We selected these two birth weight groups because previous NICHD data showed an increase in mortality of approximately 20% in infants with birth weights ≤1000 g who had received PT.(8)

Another subgroup analysis was performed to determine the contrast between the two groups. Exploratory analyses were performed of survivors and non-survivors.

We used the t test, Mann-Whitney test, or Fisher exact test to compare continu-ous variables and the chi-square test for categorical variables. A two-sided P value of <0.05 was considered statistically significant. Statistical analyses were performed with SPSS 18.0 software for Windows (SPSS Inc, Chicago, IL).

We did not plan an interim analysis on the primary outcome since these data would not be available before 18 to 24 months. An independent data and safety monitoring committee reviewed serious interim results, including death and other adverse outcomes of prematurity, and the rate of exchange transfusions after inclu-sion of 200 infants.    

Results

Between April 2007 and April 2008, 615 (66%) out of 934 eligible infants were enrolled (Fig. 1). On the whole, the B/A ratio and TSB groups were similar with regards to the baseline characteristics (Table 1).

58

934 infants assessedfor eligibility 12

121145283

10

did not meet eligibility criteriano consent by parents/guardiansconsent not sought forexcluded for other reasonswithdrawndied before randomization

615 underwentrandomization

306 assigned tobilirubin/albuminratio group

16 died beforedischarge

25 died beforedischarge

309 assigned tototal serum bilirubin group

290 survived todischarge

0 died afterdischarge

1 died afterdischarge

284 survived todischarge

290 survived to24 months

50 had unknown neuro-developmental statusat 18–24 months

240 had known neuro-developmental statusat 18–24 months

238 had known neuro-developmental statusat 18–24 months

45 had unknown neuro-developmental statusat 18–24 months

283 survived to24 months

Figure 1. Enrollment, randomization, and follow-up At follow up there were no significant baseline differences between infants with known and those with unknown neurodevelopmental status.

The BARTrial 59

Table 1. Baseline characteristics at randomization

Characteristics Bilirubin/albumin ratio group

(n=306)

Total serum bilirubin group

(n=309)

Gestational age – wk 29 ± 2 29 ± 2

Birth weight – g 1264 ± 360 1250 ± 330

Small for gestational age – no.(%) 70 (23) 73 (24)

Male sex – no. (%) 171 (56) 181 (59)

Multiple birth – no.(%) 98 (32) 104 (34)

Race or ethnicity# – no./total no.(%):

Caucasian 253/299 (83) 263/304 (85)

Mediterranean 12/299 (4) 11/304 (4)

Asian 10/299 (3) 14/304 (5)

African 15/299 (5) 13/304 (4)

Latin-American 9/299 (3) 3/304 (1)

Antenatal steroids – no./total no.(%) 257/290 (89) 246/295 (83)

Caesarean delivery – no./total no.(%) 139/304 (46) 138/297 (46)

5 min. Apgar score <3 – no./total no.(%) 5/302 (2) 6/306 (2)

Positive Coombs test – no. (%) 1 (0.3) 1 (0.3)

Maternal irregular antibodies – no./total no.(%) 8/197 (4) 5/202 (2)

wk = weeks, g = grams, no. = number, min. = minutes

Plus-minus values are means ± standard deviations. The denominator used to calculate the percent-ages of infants or mothers with a specific characteristic was the number for whom the characteris-tic was known. This number was the total number in each group, unless otherwise specified.

No significant differences were found between the two groups.

# Race or ethnicity was reported by patients’ parents or guardians or determined by the physician on reviewing the charts.

Outcome at 18 to 24 months

Table 2 shows the neurodevelopmental outcome at 18 to 24 months’ corrected age for 478 infants, i.e. 78% of the 615 randomized infants and 83% of the 573 infants who survived until neurodevelopmental testing. The primary outcome, i.e. the mean (± SD) composite motor score for the two groups was similar. The number of infants with a composite motor score of <85 was also similar: 12 (5.1%) out of 237 in the B/A ratio group versus 9 (3.7%) out of 243 in the TSB group (P=0.51). We found no differences between the two groups on the fine and gross motor scales, nor on the composite cognitive scale. Moreover, the additional use of the B/A ratio did not reduce the rate of NDI, severe or otherwise.

60

Tabl

e 2. O

utco

mes

at 18

to 24

mon

ths

Bi

lirub

in/a

lbum

in

rati

o g

roup

Tota

l ser

um

b

iliru

bin

gro

up

P va

lue

#RR (

95%

CI)

– P

va

lue

$

Prim

ary

outc

om

e

Com

posi

te m

oto

r sc

ore

100

±13

101±

120

.49

–

Fin

e m

oto

r sc

ale

11±3

12±3

0.2

8–

Gro

ss m

oto

r sc

ale

9±2

9±2

0.9

5–

Com

posi

te c

og

nit

ive s

core

101±

1210

1±11

0.6

1–

Seco

ndary

outc

om

e –

no./to

tal n

o. (

%)

Death

16/3

06

(5)

26/3

09

(8)

0.15

0.6

2 (0

.34

–1.14

) –

0.12

Sev

ere n

euro

dev

elo

pm

enta

l im

pair

ment

10/2

64

(4

)5/

269

(2)

0.2

02.

04

(0

.71–

5.8

8)

– 0

.19

Neuro

dev

elo

pm

enta

l im

pair

ment

49/2

64

(19

)59/2

69

(22

)0

.39

0.8

5 (

0.6

0–1

.19)

– 0

.33

Com

posi

te m

oto

r sc

ore

<70

4/2

37 (

2)2/

243 (

1)0

.44

2.0

5 (

0.3

8–1

1.09

) –

0.4

0

Com

posi

te m

oto

r sc

ore

<85

12/2

37 (

5)

9/2

43 (

4)

0.5

11.3

7 (0

.59

–3.18

) –

0.4

7

Com

posi

te c

og

nit

ive s

core

<70

4/2

56

(2)

0/2

64

(0

)0

.06

9.2

8 (

0.5

0–1

71)

– 0

.13

Com

posi

te c

og

nit

ive s

core

<85

10/2

56

(4

)11

/26

4 (

4)

1.00

.94

(0

.41–

2.17

) –

0.8

8

Cer

eb

ral p

als

y8/2

64

(3)

4/2

69

(1)

0.2

62.

04

(0

.62–

6.6

9)

– 0

.24

Sev

ere h

eari

ng

loss

1/26

4 (

<1)

1/26

9(<

1)1.0

1.02

(0.0

6–1

6.2

) –

0.9

9

Any

heari

ng

imp

air

ment

9/2

64

(3)

12/2

69

(5)

0.6

60

.76

(0

.33–1

.78)

– 0

.53

Sev

ere v

isua

l im

pair

ment

0/2

64

(0

)1/

269

(<1)

1.00

.34

(0

.014

–8.3

) –

0.5

1

Any

visu

al i

mp

air

ment

18/2

64

(7)

23/2

69

(9

)0

.52

0.8

0 (

0.4

4–1

.44

) –

0.4

5

#: O

utc

om

e o

f Fis

cher

exa

ct t

est

, tw

o-t

aile

d; $

: rela

tive

ris

k of

the B

/A r

ati

o g

roup

in c

om

pari

son t

o t

he T

SB

gro

up

.

The d

enom

inato

r use

d t

o c

alc

ulate

the p

erce

nta

ge o

f in

fants

wit

h a

sp

eci

fic

outc

om

e w

as

the n

um

ber

of

infa

nts

rand

om

ly a

ssig

ned

to e

ach

tre

atm

ent

gro

up

for

whom

th

at

outc

om

e w

as

know

n a

t 18

to 2

4 m

onth

s. T

his

num

ber

was

the t

ota

l num

ber

in e

ach

gro

up

, unle

ss o

ther

wis

e s

peci

fied

. The m

oto

r and

cog

nit

ive s

core

s w

ere a

ssess

ed

wit

h t

he B

SID

III (

score

s ra

ng

e f

rom

50

to 150

, wher

e 150

ind

icate

s m

ost

ad

vance

d d

evelo

pm

ent)

.

The r

ela

tive

ris

k of

each

outc

om

e w

as

calc

ulate

d for

the B

/A r

ati

o g

roup

as

com

pare

d t

o t

he T

SB

gro

up

.

In t

he B

/A r

ati

o g

roup

the m

ean (

±SD

) a

ge a

t d

eath

was

30

± 16

da

ys a

nd 10

± 7

da

ys in

the T

SB

gro

up

. Sev

ere N

DI w

as

a c

om

posi

te m

oto

r sc

ore

of

<70

, a c

om

posi

te c

og

nit

ive s

core

of

<70

, mod

erate

or

seve

re c

ereb

ral p

als

y, s

ever

e u

nila

tera

l or

bila

tera

l heari

ng

loss

, or

unila

tera

l or

bila

tera

l blin

dness

. N

euro

dev

elo

pm

enta

l im

pair

ment

was

a c

om

posi

te m

oto

r sc

ore

of

<85, a

com

posi

te c

og

nit

ive s

core

of

<85, a

ny

neuro

log

ica

l im

pair

ment, a

ny

visu

al

imp

air

ment, o

r heari

ng

imp

air

ment.

The BARTrial 61

The two groups did not differ regarding TSB levels and B/A ratios (Table 3) and the number and duration of PT sessions was similar. Two infants in the B/A ratio group received an exchange transfusion against none in the TSB group.

Table 3 shows secondary outcomes during NICU hospitalization. Fewer infants allocated to the B/A ratio group died during the first two weeks. Although not sta-tistically significant, the overall mortality rate until the age of 18 to 24 months also appeared somewhat lower in the B/A ratio versus the TSB group (RR, 0.64; 95% CI, 0.35 to 1.19; P=0.12). This prompted us to analyze mortality in the subgroups in more detail.

62

Tabl

e 3. B

iliru

bin-

rela

ted

cour

se an

d se

cond

ary o

utco

mes

dur

ing h

ospi

taliz

atio

n

Bilir

ub

in/a

lbum

in

rati

o g

roup

Tota

l ser

um

b

iliru

bin

gro

up

P va

lue

Neonata

l cours

en

=30

6n

=30

9

Peak

tota

l ser

um

bili

rub

in –

µm

ol/

L17

9±4

418

1±4

60

.66

Mean t

ota

l ser

um

bili

rub

in –

µm

ol/

L12

4±3

512

7±36

0.5

1

Troug

h a

lbum

in –

g/L

25.5

±5.2

25.4

±5.2

0.7

6

Mean a

lbum

in –

g/L

29.1±

4.8

29.4

±5.4

0.5

4

Peak

B/A

-rati

o –

µm

ol/g

6.2

±1.4

6.3

±1.7

0.4

1

Mean B

/A-r

ati

o –

µm

ol/g

4.4

±1.1

4.4

±1.2

0.3

6

Photo

ther

ap

y –

no. (

%)

270

(88%

)26

8 (

87%

)0

.63

Dura

tion o

f p

hoto

ther

ap

y –

hrs

77±5

171±

48

0.12

Exch

ang

e tra

nsf

usi

ons

– no.

20

0.2

5

Seco

ndary

outo

me –

no./to

tal n

o (

%)

P va

lue

#RR (

95%

CI)

– P

va

lue

$

Death

befo

re d

ay

154

/30

6 (

1.3)

17/3

09

(5.5

)0

.00

63*

0.2

4 (

0.0

8–0

.70

) –

0.0

09

*

Death

befo

re d

isch

arg

e16

/30

6 (

5.2

)25

/30

9 (

8.1)

0.2

00

.64

(0

.35–1

.19)

– 0

.16

Sep

sis

91/

30

4 (

30

)88/3

08 (

29)

0.7

21.0

5 (

0.8

2-.13

4)

– 0

.71

Menin

git

is2/

30

0 (

0.7

)1/

30

6 (

0.3

)0

.25

2.0

(0

.19–2

2) –

0.5

6

Intr

ave

ntr

icul

ar

hem

orr

ha

ge, g

rad

es

3 o

r 4

15/3

04

(5)

19/3

07

(6)

0.6

00

.80

(0

.41–

1.54

) –

0.5

0

Pate

nt

duct

us

art

erio

sus

92/

30

6 (

30

)9

8/3

09

(32)

0.8

70

.95 (

0.7

5–1

.20

) –

0.6

6

treate

d w

ith s

urg

ica

l lig

ati

on

15/3

06

(5)

17/3

09

(6

)0

.86

0.8

9 (

0.4

5–1

.75)

– 0

.74

Necr

oti

zing

ente

roco

litis

28/3

06

(9

)31/

30

9 (

10)

0.7

90

.91 (0

.56

–1.4

8)

– 0

.71

treate

d s

urg

ica

lly11

/30

6 (

4)

12/3

09

(4

)1.0

0.9

3 (

0.4

1–2.

10)

– 0

.85

Peri

ventr

icul

ar

leuco

ma

laci

a, g

rad

es

3 o

r 4

2/30

4 (

0.7

)1/

30

7 (0

.3)

0.2

52.

02

(0.18

–22)

– 0

.57

Bronch

op

ulm

onary

dys

pla

sia a

t 36

wks

’ p

ost

menst

rua

l ag

e26

/30

6 (

9)

32/

30

9 (

10)

0.4

90

.52

(0.5

0–1

.34

) –

0.4

3

Ret

inop

ath

y of

pre

matu

rity

gra

de 3

or

hig

her

6/3

04

(2)

4/3

09

(1)

0.5

41.5

2 (0

.43–5

.35)

– 0

.51

ALG

O r

efe

r14

/26

9 (

5)

16/2

70 (

6)

0.8

50

.88 (

0.4

4–1

.76

) –

0.7

2

#: O

utc

om

e o

f Fis

cher

exa

ct t

est

, tw

o-t

aile

d; $

: rela

tive

ris

k of

the B

/A r

ati

o g

roup

in c

om

pari

son t

o t

he T

SB

gro

up

.

Plus-

min

us

valu

es

are

means

± st

and

ard

dev

iati

ons.

The d

enom

inato

r use

d t

o c

alc

ulate

the p

erce

nta

ge o

f in

fants

wit

h a

sp

eci

fic

outc

om

e w

as

the

num

ber

for

whom

the v

ari

ab

le w

as

know

n. T

his

num

ber

was

the t

ota

l num

ber

in e

ach

gro

up

, unle

ss o

ther

wis

e s

peci

fied

.

The BARTrial 63

Subgroup analyses

Table 4 shows the results of the subgroup analyses according to birth weights. Among the 162 infants with birth weights ≤1000 g the additional use of the B/A ratio did not reduce mortality. Among the 453 infants with birth weights of >1000 g, mortality in the B/A ratio group was significantly lower than in the TSB group (RR, 0.30; 95%CI: 0.10–0.92, P=0.03).

64

Outc

om

es

B/A

rati

o g

roup

TS

B g

roup

P va

lue

Com

posi

te m

oto

r sc

ore

≤1

00

0 g

100

±11 (5

6)

100

±11 (5

1)0

.89

>10

00

g10

1±13

(16

5)

102±

13 (

166

)0

.49

Gro

ss m

oto

r sc

ale

≤1

00

0 g

9±3

(55)

9±2

(53)

0.8

3

>10

00

g9

±2 (

166

)9

±2 (

166

)0

.70

Fin

e m

oto

r sc

ale

≤1

00

0 g

11±2

(57

)11

±3 (

53)

0.4

0

>10

00

g11

.34

±2.6

0 (

177)

12±3

(17

6)

0.4

4

Com

posi

te c

og

nit

ive s

core

≤1

00

0 g

101±

11 (

58)

100

±11 (5

7)0

.90

>10

00

g10

1±12

(18

2)10

1±11

(18

1)0

.61

Com

posi

te m

oto

r sc

ore

<70

≤1

00

0 g

0/6

80

/64

1.0

0

>10

00

g4

/16

9 (

2.4

%)

2/17

9 (

1.1%

)0

.44

Com

posi

te m

oto

r sc

ore

<85

≤1

00

0 g

2/6

8 (

2.9

%)

1/6

4 (

1.6%

)1.0

>10

00

g10

/16

9 (

5.9

%)

8/1

79 (

4.5

%)

0.6

3

Com

posi

te c

og

nit

ive s

core

<70

≤1

00

0 g

0/7

00

/70

1.00

>10

00

g4

/186

(2.

2%)

0/1

94

0.0

57

Com

posi

te c

og

nit

ive s

core

<85

≤1

00

0 g

1/70

(1.4

%)

2/70

(2.

9%

)1.0

0

>10

00

g9/1

86

(4

.8%

)9/1

94

(4

.6%

)1.0

0

Death

≤1

00

0 g

12/7

8 (

15.4

%)

13/8

4 (

15.5

%)

1.00

>10

00

g4

/228

(1.8

%)

13/2

25 (

5.8

%)

0.0

27*

Sev

ere N

DI #

≤1

00

0 g

2/71 (2

.8%

)1/

72 (

1.4%

)1.0

>10

00

g8/1

93 (

4.1%

)4

/197

(2.0

%)

0.2

6

Death

or

seve

re N

DI

≤1

00

0 g

14/7

1 (1

9.7

%)

14/7

2 (1

9.4

%)

1.00

>10

00

g12

/19

3 (

6.2

%)

17/1

97

(8.6

%)

0.4

4

ND

I

≤1

00

0 g

13/7

1 (1

8.3

%)

12/7

2 (1

6.7

%)

0.8

3

>10

00

g36

/19

3 (

18.7

%)

47/

197

(23.9

%)

0.2

2

Death

or

ND

I

≤1

00

0 g

25/7

1 (3

5.2

%)

25/7

2 (3

4.7

%)

1.00

>10

00

g4

0/1

93 (

20.7

%)

60

/197

(30

.5%

)0

.037

*

Cer

eb

ral p

als

y

≤1

00

0 g

1/71 (1

.4%

)0

/72

0.5

0

>10

00

g7/

193 (

3.6

%)

4/1

97

(2.2

%)

0.3

8

Sev

ere h

eari

ng

loss

≤1

00

0 g

1/71 (1

.4%

)0

/72

0.5

0

>10

00

g0

/19

31/

197

(0.5

%)

1.00

Any

heari

ng

imp

air

ment

≤1

00

0 g

2/71 (2

.8%

)2/

72 (

2.8%

)1.0

0

>10

00

g7/

193 (

3.6

%)

10/1

97

(5.1%

)0

.62

Tabl

e 4. S

tratifi

ed an

alys

is of

out

com

es at

18 to

24 m

onth

s, ac

cord

ing t

o bi

rth

wei

ghts

Outc

om

es

B/A

rati

o g

roup

TS

B g

roup

P va

lue

Sev

ere v

isua

l im

pair

ment

≤1

00

0 g

0/7

11/

72 (

1.4%

)1.0

0

>10

00

g0

/ 19

30

/197

1.00

Any

visu

al i

mp

air

ment

≤1

00

0 g

3/7

1 (4

.2%

)5/

72 (

6.9

%)

0.7

2

>10

00

g15

/19

3 (

7.8%

)18

/197

(9.1%

)0

.72

Sep

sis

≤1

00

0 g

43/7

8 (

55%

)31/

84

(37

%)

0.0

27*

>10

00

g4

8/ 2

26 (

21%

)57

/224

(25

%)

0.3

2

IVH

all

gra

des

≤1

00

0 g

26/7

8 (

33%

)18

/84

(21

%)

0.11

>10

00

g4

1/22

6 (

18.1%

)50

/223

(22

%)

0.2

9

IVH

> g

rad

e 2

≤1

00

0 g

10/7

8 (

12.8

%)

10/8

4 (

11.9

%)

1.00

>10

00

g5/

226

(2.

2%)

9/2

23 (

4.0

%)

0.2

9

PDA

≤1

00

0 g

46

/78 (

59

%)

44

/84

(52%

)0

.43

>10

00

g4

6/2

28 (

20%

)54

/225

(24

%)

0.3

7

PDA

surg

ery

≤1

00

0 g

10/7

8 (

12.8

%)

10/8

4 (

11.9

%)

1.00

>10

00

g5/

228 (

2.2%

)7/

225 (

3.1%

)0

.57

NEC

all

≤1

00

0 g

18/7

8 (

23%

)12

/84

(14

.3%

)0

.16

>10

00

g10

/228

(4

.4%

)19

/225

(8.4

%)

0.0

89

NEC

surg

ery

≤1

00

0 g

8/7

8 (

10.3

%)

5/84

(6

.0%

)0

.39

>10

00

g3/2

28 (

1.3%

)7/

225 (

3.1%

)0

.22

CLD

≤1

00

0 g

30

/78 (

38%

)34

/84

(4

0%

)0

.87

>10

00

g19

/228

(8.3

%)

19/2

25 (

8.4

%)

1.00

BPD

≤1

00

0 g

18/7

8 (

23%

)20

/84

(24

%)

1.00

>10

00

g8/2

28 (

3.5

%)

12/2

25 (

5.3

%)

0.3

7

PVL

all

≤1

00

0 g

19/7

8 (

24%

)20

/84

(24

%)

1.00

>10

00

g6

2/22

6 (

27%

)6

2/22

3 (

28%

)1.0

0

PVL

> g

rad

e 2

≤1

00

0 g

1/78

(1.3

%)

1/84

(1.2

%)

1.00

>10

00

g1/

226

(0

.4%

)0

/223

1.00

RO

P a

ll

≤1

00

0 g

21/7

8 (

27%

)23

/84

(27

%)

1.00

>10

00

g12

/226

(5.3

%)

7/22

3 (

3.1%

)0

.35

RO

P >2

or

Plus

≤1

00

0 g

3/7

8 (

3.8

%)

3/8

4 (

3.6

%)

1.0

>10

00

g3/2

26 (

1.3%

)1/

223 (

0.4

%)

0.6

2

ALG

O r

efe

r

≤1

00

0 g

5/6

2 (8

.1%)

3/6

9 (

4.3

%)

0.4

8

>10

00

g9/2

07

(4.3

%)

13/2

01 (6

.5%

)0

.39

The BARTrial 65

Outc

om

es

B/A

rati

o g

roup

TS

B g

roup

P va

lue

Com

posi

te m

oto

r sc

ore

≤1

00

0 g

100

±11 (5

6)

100

±11 (5

1)0

.89

>10

00

g10

1±13

(16

5)

102±

13 (

166

)0

.49

Gro

ss m

oto

r sc

ale

≤1

00

0 g

9±3

(55)

9±2

(53)

0.8

3

>10

00

g9

±2 (

166

)9

±2 (

166

)0

.70

Fin

e m

oto

r sc

ale

≤1

00

0 g

11±2

(57

)11

±3 (

53)

0.4

0

>10

00

g11

.34

±2.6

0 (

177)

12±3

(17

6)

0.4

4

Com

posi

te c

og

nit

ive s

core

≤1

00

0 g

101±

11 (

58)

100

±11 (5

7)0

.90

>10

00

g10

1±12

(18

2)10

1±11

(18

1)0

.61

Com

posi

te m

oto

r sc

ore

<70

≤1

00

0 g

0/6

80

/64

1.0

0

>10

00

g4

/16

9 (

2.4

%)

2/17

9 (

1.1%

)0

.44

Com

posi

te m

oto

r sc

ore

<85

≤1

00

0 g

2/6

8 (

2.9

%)

1/6

4 (

1.6%

)1.0

>10

00

g10

/16

9 (

5.9

%)

8/1

79 (

4.5

%)

0.6

3

Com

posi

te c

og

nit

ive s

core

<70

≤1

00

0 g

0/7

00

/70

1.00

>10

00

g4

/186

(2.

2%)

0/1

94

0.0

57

Com

posi

te c

og

nit

ive s

core

<85

≤1

00

0 g

1/70

(1.4

%)

2/70

(2.

9%

)1.0

0

>10

00

g9/1

86

(4

.8%

)9/1

94

(4

.6%

)1.0

0

Death

≤1

00

0 g

12/7

8 (

15.4

%)

13/8

4 (

15.5

%)

1.00

>10

00

g4

/228

(1.8

%)

13/2

25 (

5.8

%)

0.0

27*

Sev

ere N

DI #

≤1

00

0 g

2/71 (2

.8%

)1/

72 (

1.4%

)1.0

>10

00

g8/1

93 (

4.1%

)4

/197

(2.0

%)

0.2

6

Death

or

seve

re N

DI

≤1

00

0 g

14/7

1 (1

9.7

%)

14/7

2 (1

9.4

%)

1.00

>10

00

g12

/19

3 (

6.2

%)

17/1

97

(8.6

%)

0.4

4

ND

I

≤1

00

0 g

13/7

1 (1

8.3

%)

12/7

2 (1

6.7

%)

0.8

3

>10

00

g36

/19

3 (

18.7

%)

47/

197

(23.9

%)

0.2

2

Death

or

ND

I

≤1

00

0 g

25/7

1 (3

5.2

%)

25/7

2 (3

4.7

%)

1.00

>10

00

g4

0/1

93 (

20.7

%)

60

/197

(30

.5%

)0

.037

*

Cer

eb

ral p

als

y

≤1

00

0 g

1/71 (1

.4%

)0

/72

0.5

0

>10

00

g7/

193 (

3.6

%)

4/1

97

(2.2

%)

0.3

8

Sev

ere h

eari

ng

loss

≤1

00

0 g

1/71 (1

.4%

)0

/72

0.5

0

>10

00

g0

/19

31/

197

(0.5

%)

1.00

Any

heari

ng

imp

air

ment

≤1

00

0 g

2/71 (2

.8%

)2/

72 (

2.8%

)1.0

0

>10

00

g7 /

193 (

3.6

%)

10/1

97

(5.1%

)0

.62

*: O

utc

om

e o

f Fis

cher

exa

ct t

est

, tw

o-t

aile

d

The d

enom

inato

r use

d t

o c

alc

ulate

the p

erce

nta

ge o

f in

fants

wit

h a

sp

eci

fic

outc

om

e w

as

the n

um

ber

of

infa

nts

rand

om

ly a

ssig

ned

to e

ach

gro

up

, and

cla

s-si

fied

in e

ach

bir

th w

eig

ht

sub

gro

up

, for

whom

the o

utc

om

e w

as

know

n. g

= g

ram

s

# S

ever

e N

DI i

s a c

om

posi

te m

oto

r sc

ore

of

<70

or

a c

om

posi

te c

og

nit

ive s

core

of

<70

, mod

erate

or

seve

re c

ereb

ral p

als

y, s

ever

e b

ilate

ral h

eari

ng

loss

or

bila

tera

l blin

dness

.

Neuro

dev

elo

pm

enta

l im

pair

ment

(ND

I) is

a c

om

posi

te m

oto

r sc

ore

of

<85 o

r a c

om

posi

te c

og

nit

ive s

core

of

<85, a

ny

neuro

log

ica

l im

pair

ment, a

ny

visu

al

imp

air

ment, o

r any

heari

ng

imp

air

ment.

Outc

om

es

B/A

rati

o g

roup

TS

B g

roup

P va

lue

Sev

ere v

isua

l im

pair

ment

≤1

00

0 g

0/7

11/

72 (

1.4%

)1.0

0

>10

00

g0

/19

30

/197

1.00

Any

visu

al i

mp

air

ment

≤1

00

0 g

3/7

1 (4

.2%

)5/

72 (

6.9

%)

0.7

2

>10

00

g15

/19

3 (

7.8%

)18

/197

(9.1%

)0

.72

Sep

sis

≤1

00

0 g

43/7

8 (

55%

)31/

84

(37

%)

0.0

27*

>10

00

g4

8/2

26 (

21%

)57

/224

(25

%)

0.3

2

IVH

all

gra

des

≤1

00

0 g

26/7

8 (

33%

)18

/84

(21

%)

0.11

>10

00

g4

1/22

6 (

18.1%

)50

/223

(22

%)

0.2

9

IVH

> g

rad

e 2

≤1

00

0 g

10/7

8 (

12.8

%)

10/8

4 (

11.9

%)

1.00

>10

00

g5/

226

(2.

2%)

9/2

23 (

4.0

%)

0.2

9

PDA

≤1

00

0 g

46

/78 (

59

%)

44

/84

(52%

)0

.43

>10

00

g4

6/2

28 (

20%

)54

/225

(24

%)

0.3

7

PDA

surg

ery

≤1

00

0 g

10/7

8 (

12.8

%)

10/8

4 (

11.9

%)

1.00

>10

00

g5/

228 (

2.2%

)7/

225 (

3.1%

)0

.57

NEC

all

≤1

00

0 g

18/7

8 (

23%

)12

/84

(14

.3%

)0

.16

>10

00

g10

/228

(4

.4%

)19

/225

(8.4

%)

0.0

89

NEC

surg

ery

≤1

00

0 g

8/7

8 (

10.3

%)

5/84

(6

.0%

)0

.39

>10

00

g3/2

28 (

1.3%

)7/

225 (

3.1%

)0

.22

CLD

≤1

00

0 g

30

/78 (

38%

)34

/84

(4

0%

)0

.87

>10

00

g19

/228

(8.3

%)

19/2

25 (

8.4

%)

1.00

BPD

≤1

00

0 g

18/7

8 (

23%

)20

/84

(24

%)

1.00

>10

00

g8/2

28 (

3.5

%)

12/2

25 (

5.3

%)

0.3

7

PVL

all

≤1

00

0 g

19/7

8 (

24%

)20

/84

(24

%)

1.00

>10

00

g6

2/22

6 (

27%

)6

2/22

3 (

28%

)1.0

0

PVL

> g

rad

e 2

≤1

00

0 g

1/78

(1.3

%)

1/84

(1.2

%)

1.00

>10

00

g1/

226

(0

.4%

)0

/223

1.00

RO

P a

ll

≤1

00

0 g

21/7

8 (

27%

)23

/84

(27

%)

1.00

>10

00

g12

/226

(5.3

%)

7/22

3 (

3.1%

)0

.35

RO

P >2

or

Plus

≤1

00

0 g

3/7

8 (

3.8

%)

3/8

4 (

3.6

%)

1.0

>10

00

g3/2

26 (

1.3%

)1/

223 (

0.4

%)

0.6

2

ALG

O r

efe

r

≤1

00

0 g

5 /6

2 (8

.1%)

3/6

9 (

4.3

%)

0.4

8

>10

00

g9/2

07

(4.3

%)

13/2

01 (6

.5%

)0

.39

66

The composite motor score was not affected by treatment allocation in the two birth weight groups. The additional use of the B/A ratio resulted in an increase in the percentage of infants with a composite cognitive score of <70 in the subgroup with birth weights >1000 g (2.2% for the B/A ratio versus 0% for the TSB group, P=0.057). The reduction of approximately 10% in the composite outcome of death or NDI in infants with birth weights of >1000 g allocated to the B/A ratio group was, therefore, attributed to the reduction in mortality. Correspondingly, fewer adverse secondary outcomes, e.g. sepsis, intraventricular hemorrhage, patent ductus arte-riosus, bronchopulmonary dysplasia, and necrotizing enterocolitis were observed in this group, although the reduction in none of these morbidities was statistically significant. Adverse secondary outcomes tended to occur more frequently in the B/A ratio subgroup of infants weighing ≤1000 g. The 18% increase in the incidence of sepsis was statistically significant  (RR, 1.5; 95% CI: 1.1–2.1, P=0.027).

Supplementary Table 5 shows the results of the subgroup analysis regarding treat-ment based on B/A ratio. In the B/A ratio group, 30% of the infants received PT at least once on the basis of the B/A ratio crossing its threshold and not on the basis of the TSB level. In retrospect, 25% of the infants in the TSB group had crossed the B/A ratio and not the TSB threshold, but were by group definition not treated with PT. Developmental outcomes did not differ between the two groups. We found a tendency towards fewer mortalities in the B/A ratio group: 7 (7.6%) out of 92 ver-sus 13 (16.9%) out of 77 (RR, 0.45; 95% CI, 0.19 to 1.07, P=0.09). Mortality was not reduced in the B/A ratio subgroup of infants who were not treated on the basis of the B/A ratio (RR, 0.75; 95% CI, 0.33 to 1.72, P=0.52). Despite comparable TSB and albumin levels, peak and mean B/A ratios were significantly lower in the B/A ratio group, without any effect on PT duration.

Relationship of bilirubin and outcome

Peak B/A ratios during the first ten days were >12% higher in the non-survivors compared to the survivors (P=0.004); peak TSB levels and trough (±SD) albumin levels were significantly lower in the non-survivors (supplementary Table 6).

The peak B/A ratios and TSB levels of the survivors without hearing loss or with a composite motor score ≥85 were comparable to the infants with any amount of hearing loss or a composite motor score of <85.

The BARTrial 67

Supplementary Table 5. Baseline characteristics, outcomes, and bilirubin-related course of infants who were treated at least once on the basis of their B/A ratio and not on the basis of TSB levels versus control infants who at least once had a B/A ratio that exceeded the B/A ratio threshold, but who were not treated accordingly.

B/A ratio TSB P value

Baseline characteristics n=92 n=77

birth weight (g) 1258±364 1232±313 0.62

gestational age (wks) 29±1.8 29±1.8 0.81

Outcomes

Mortality (n/N (%)) 7/92 (7.6%) 13/77 (16.9%) 0.09

Motor composite score 101±12 103±12 0.51

– Fine motor score 12±2 12±3 0.32

– Gross motor score 9±2 9±2 0.77

Cognitive composite score 102±12 102±10 0.82

NDI severe (n/N (%)) 4/81 (4.9%) 0/68 0.38

NDI mild 15/81 (18.5%) 11/68 (16.2%) 0.83

NDI mild + severe 19/81 (23%) 11/68 (16.2%) 0.42

Death or severe NDI 11/81 (13.6%) 14/68 (21%) 0.28

TSB, B/A and PT

TSB max (µmol/L) 174±42 186±48 0.11

TSB mean (µmol/L) 121±36 127±35 0.26

B/A ratio max (µmol/g) 7.05±1.37 7.72±1.83 0.01*

B/A mean (µmol/g) 4.88±1.09 5.27±1.25 0.03*

Alb trough (g/L) 20.7±4.0 20.2±4.3 0.36

Alb mean (g/L) 24.8± 4.3 24.4±4.1 0.51

PT duration (hrs) 87±60 77±55 0.29

Means and SD, or n/N (%). P values are the outcomes of either the t test or Fisher exact test. * P<0.05.

B/A ratio group: n=92, 7 died, 11 lost to follow-up.

TSB group: n=77, 13 died, 9 lost to follow-up.

68

Supplementary Table 6. Bilirubin-related values of the different subgroups

a) Survivors versus non-survivors

Survivors (n=572) Non-survivors (n=42)# P value

TSB max (µmol/L) 181±44 166±47 0.036*

TSB mean (µmol/L) 127±35 111±35 0.005*

B/A ratio max (µmol/g) 6.19±1.48 6.91±2.38 0.004*

B/A mean (µmol/g) 4.37±1.12 4.65±1.56 0.145

Albumin trough (g/L) 25.8±4.9 20.5±5.8 0.000*

Albumin mean (g/L) 29.5± 4.5 25.6±9.3 0.011*

PT duration (hrs) 73±50 84±53 0.228

#: One patient’s TSB values are missing; *: Outcome of t test, two-tailed.

b) Neurodevelopmentally non-impaired versus NDI survivors

Non-impaired (n=383) NDI (n=107) P value

TSB max (µmol/L) 181±44 182±45 0.88

TSB mean (µmol/L) 128±35 129±35 0.67

B/A ratio max (µmol/g) 6.16±1.44 6.30±1.52 0.38

B/A mean (µmol/g) 4.34±1.06 4.50±1.13 0.18

Albumin trough (g/L) 26.1±5.1 25.4±4.8 0.19

Albumin mean (g/L) 29.7±4.7 29.1±4.5 0.18

PT duration (hrs) 74±48 80±60 0.25

c) Survivors with a composite motor scores ≥85 versus <85

PDI ≥85 (n=480) PDI <85 (n=21) P value

TSB max (µmol/L) 181±45 170±39 0.29

TSB mean (µmol/L) 127±35 122±28 0.59

B/A ratio max (µmol/g) 6.20±1.47 5.97±1.61 0.48

B/A mean (µmol/g) 4.36±1.08 4.34±1.02 0.95

Albumin trough (g/L) 25.8±5.1 25.0±4.3 0.50

Albumin mean (g/L) 29.5±4.7 29.0±4.11 0.66

PT duration (hrs) 77±50 85±86 0.56

PDI: Psychomotor Developmental Index

The BARTrial 69

d) Survivors with normal hearing versus any hearing impairment

Normal hearing (n=464)

Any hearing impairment (n=19)

P value

TSB max (µmol/L) 181±43 194±55 0.24

TSB mean (µmol/L) 128±34 136±37 0.35

B/A ratio max (µmol/g) 6.18±1.42 6.79±2.11 0.07

B/A mean (µmol/g) 4.37±1.06 4.77±1.39 0.12

Albumin trough (g/L) 26±5 26.4±9.98 0.72

Albumin mean (g/L) 29.6±4.6 29.4±5.74 0.85

PT duration (hrs) 75±52 66±29 0.47

Discussion

In this multicenter trial we found no difference in the rate of the composite motor score at 18 to 24 months’ corrected age between the infants randomly allocated to receive PT and/or exchange transfusion based on the additional use of the B/A ratio and the infants allocated to receive treatment based exclusively on TSB levels. Nor did we find differences in other clinical outcomes including death. In pre-specified subgroup analyses mortality among the infants with birth weights of >1000 g was significantly lower in those allocated to the B/A ratio group, while mortality in both subgroups of infants weighing ≤1000 g was similar. Furthermore, we observed no significant differences in any of the bilirubin parameters between the survivors with NDI, severe or otherwise, or hearing loss and those without NDI or hearing loss.

The absence of any effect of the additional use of the B/A ratio on infants’ neuro-logical development including hearing loss or mortality may be explained, at least partly, by similar TSB levels, similar B/A ratios, and no significant differences in PT treatment. In preterms, each of these variables is putatively involved in the risk of bilirubin-associated NDI or death.(8,9) Considering the occurrence of bilirubin encephalopathy in preterms with low TSB levels the risk of developing bilirubin neurotoxicity is not determined by TSB alone.(10–12) Only free bilirubin, and per-haps the B/A ratio, are more closely associated with bilirubin neurotoxicity.(2,13) Nevertheless, a correlation between peak TSB levels and NDI or death was found in infants with extremely low birth weights (ELBW).(14,15) This finding is in agree-ment with the reduction in overall NDI at lower TSB levels in ELBW infants in the ‘aggressive’ PT part of a recent randomized controlled trial (RCT).(9) In this as well as in another RCT a trend was found toward increased mortality in ELBW infants who had received more PT.(8,16)

70

In our RCT fewer deaths in the B/A ratio subgroup of infants with birth weights of >1000 g was not associated with less PT (data not shown) or with any significant differences in neonatal morbidities. The observed difference in mortality seemed biologically unrelated to the management of hyperbilirubinemia and could thus be considered a chance occurrence. Nevertheless, the mortality rate in this subgroup, 1.8%, was also less than the historical figure of approximately 6% kept by the Dutch Perinatal Registry for infants who had also been treated according to their TSB lev-els. Furthermore, the reduction in mortality between the two categories of our trial appeared to be consistent. We found a trend towards fewer deaths in the subgroup allocated to the B/A ratio group who had been treated at least once on the basis of their B/A ratios (-9.3%, P=0.09) and who had statistically significant lower (mean) peak B/A ratios.

We did not find significant differences in causes of death between the two treat-ment and birth weight categories, although there was a trend towards fewer infants who had necrotizing enterocolitis in the B/A ratio subgroup weighing >1000 g at birth. Infants in this subgroup lived significantly longer compared to infants allocated to the TSB category (30 ± 16 versus 10 ± 7 days, P=0.003). Overall, peak B/A ratios were significantly higher in the non-survivors, whereas TSB and albumin levels were actually lower. We do not know whether the infants who died had a reduced anti-oxidant capacity due to lower albumin and bilirubin levels. This finding supported our original hypothesis that the additional use of the B/A ratio is a valuable parameter in hyperbilirubinemia management in preterms. It is tempting to speculate that it may be more appropriately to treat infants with PT when bilirubin is more harmful and the infant brain is more vulnerable to bilirubin neurotoxicity, i.e. when albumin levels are low. The pathophysiological conditions may involve respiratory and circula-tory failure and sepsis, as was previously described.(17) Oh et al. found that higher Bf levels are associated with a higher risk of death or adverse neurodevelopmental outcome regardless of clinical status. This finding also supports the potential benefits of using the B/A ratio in addition to the TSB level. It is a more useful indicator of the Bf level than using the TSB level alone.(18,19) Using the B/A ratio in addition to TSB levels may lead to reducing early mortality by a mechanism that is as yet unknown. In a trial performed more than half a century ago similar inexplicable observations were reported. It showed that a specific antibiotic treatment increases the risk of kernicterus and through basic scientific investigations this led to new insights in the mechanism of kernicterus.(20) Our unexpected finding that use of the B/A ratio in the management of preterms with birth weights of >1000 g reduces mortality needs to be confirmed by a larger study.

The BARTrial 71

How can our results be translated into clinical practice? Incorporating B/A ratios in management guidelines does not seem an essential parameter to improve neuro-developmental outcome, but it may reduce mortality in infants weighing >1000 g at birth. More specifically, the additional use of pre-defined B/A ratios did not allow for earlier or more intensive treatment and lower TSB levels. This was, at least in part, what we expected and is related to its add-on use. We found no evidence for safe B/A ratio thresholds that would justify the sole use of B/A ratio rather than TSB thresholds. The contrast between the two treatment groups was thus diluted. Higher pre-specified albumin levels and lower B/A ratios would have resulted in more contrast, i.e. earlier treatment based on the B/A ratio. Actually, the mean (±SD) albumin levels were 27.8 (±5.5) versus 30.8 (±4) for infants with birth weights of <1250 g (n=319) versus ≥1250 g (n=291). Based on the relatively low differences with pre-specified albumin levels, we anticipated a rather small contrast. Another explanation for the absence of an effect of the additional use of the B/A ratio on neurodevelopmental outcome could be the relatively low treatment thresholds: the majority of preterms received PT. To date, the applied thresholds were consensus–based. The results of our RCT provided some evidence of the benefits without causing any harm. We demonstrated that the additional use of the B/A ratio does not result in harmful PT.

Conclusion

In preterm infants of 32 weeks’ gestation or less with hyperbilirubinemia we found no significant effect of the additional use of the B/A ratio compared to TSB-based treatment on the composite motor score at 18 to 24 months’ corrected age. Even though the additional use of the B/A ratio in the management of preterms with birth weights of >1000 g reduced mortality significantly, this finding needs to be confirmed in a larger study.

72

Acknowledgements

We greatly appreciate the help of Titia Brantsma-van Wulfften Palthe in correcting the English manuscript.

This project is funded by a grant from:

The BARTrial Study group:University Medical Center Groningen, Beatrix Children’s Hospital:

C.V. Hulzebos, P.H. Dijk, D.E. van Imhoff, A.F. Bos, S.A.J. Ruiter, K.N.J.A. Van Braeckel, P.F.M. Krabbe, E.H.Quik;

Leiden University Medical Center: E. Lopriore;

Amsterdam Medical Center: L. van Toledo-Eppinga, D.H.G.M. Nuytemans, A.G. van Wassenaer-Leemhuis;

University Medical Center Utrecht: M.J.N. Benders, K.K.M. Korbeeck;

Isala Clinics Zwolle: R.A. van Lingen, L.J.M. Groot Jebbink;

University Medical Center St. Radboud Nijmegen: K.D. Liem, P. Mansvelt;

Maxima Medical Center Veldhoven: J. Buijs;

Erasmus Medical Center Rotterdam: P. Govaert, I. van Vliet;

University Medical Center Maastricht: A.L.M. Mulder, C. Wolfs;

VU Medical Center Amsterdam: W.P.F. Fetter, A.R.C. Laarman; and

Child Health Evaluative Sciences, The Hospital for Sick Children Research Institute, University of Toronto, Canada: M. Offringa.

The BARTrial 73

References

1. Wennberg RP, Ahlfors CE, Bhutani VK, Johnson LH, Shapiro SM. Toward understanding kernicterus: a challenge to improve the management of jaundiced newborns. Pediatrics 2006 Feb;117(2):474–485.

2. McDonagh A, Maisels MJ. Bilirubin Unbound: Deja Vu All Over Again? Pediatrics 2006 02/01;117(2):523–525.

3. Hulzebos CV, van Imhoff DE, Bos AF, Ahlfors CE, Verkade HJ, Dijk PH. Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity in premature infants. Arch Dis Child Fetal Neonatal Ed 2008 Sep;93(5):F384–8.

4. van Imhoff DE, Dijk PH, Weykamp CW, Cobbaert CM, Hulzebos CV, On behalf of the BARTrial Study Group. Measurements of neonatal bilirubin and albumin concentrations: a need for improvement and quality control. Eur J Pediatr 2011 Jan 7.

5. van Imhoff DE, Dijk PH, Hulzebos CV, on behalf of the BARTrial studygroup of the Netherlands Neonatal Research Network. Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a national guideline. Early Hum Dev 2011 May 27.

6. http://www.neonatologiestudies.nl/bartrial/docs/docs.asp.

7. Surveillance of Cerebral Palsy in Europe. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Surveillance of Cerebral Palsy in Europe (SCPE). Dev Med Child Neurol 2000 Dec;42(12):816–824.

8. Lipsitz PJ, Gartner LM, Bryla DA. Neonatal and infant mortality in relation to phototherapy. Pediatrics 1985 02;75(2):422–426.

9. Morris BH, Oh W, Tyson JE, Stevenson DK, Phelps DL, O’Shea TM, et al. Aggressive vs. conservative phototherapy for infants with extremely low birth weight. N Engl J Med 2008 Oct 30;359(18):1885–1896.

10. Gartner LM, Snyder RN, Chabon RS, Bernstein J. Kernicterus: high incidence in premature infants with low serum bilirubin concentrations. Pediatrics 1970 06;45(6):906–917.

11. Govaert P, Lequin M, Swarte R, Robben S, De Coo R, Weisglas-Kuperus N, et al. Changes in globus pallidus with (pre)term kernicterus. Pediatrics 2003 12;112(1098–4275; 6):1256–1263.

12. Moll M, Goeslz R, Naegele T, Wilke M, Poets CF. Are recommended phototherapy thresholds safe enough for extremely low birth weight (ELBW) infants? A report on 2 ELBW infants with kernicterus despite only moderate hyperbilirubinemia. Neonatology 2011;99(2):90–94.

13. Hulzebos CV, van Imhoff DE, Bos AF, Ahlfors CE, Verkade HJ, Dijk PH. Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity in premature infants. Arch Dis Child Fetal Neonatal Ed 2008 09;93(5):F384-F388.

14. Oh W, Tyson JE, Fanaroff AA, Vohr BR, Perritt R, Stoll BJ, et al. Association Between Peak Serum Bilirubin and Neurodevelopmental Outcomes in Extremely Low Birth Weight Infants. Pediatrics 2003 10/01;112(4):773–779.

74

15. Mazeiras G, Roze JC, Ancel PY, Caillaux G, Frondas-Chauty A, Denizot S, et al. Hyperbilirubinemia and neurodevelopmental outcome of very low birthweight infants: results from the LIFT cohort. PLoS One 2012;7(1):e30900.

16. National Institute of Child Health and Human Development randomized, controlled trials of phototherapy for neonatal hyperbilirubinemia. Pediatrics 1985 Feb;75(2 Pt 2):385–441.

17. Oh W, Stevenson DK, Tyson JE, Morris BH, Ahlfors CE, Bender GJ, et al. Influence of clinical status on the association between plasma total and unbound bilirubin and death or adverse neurodevelopmental outcomes in extremely low birth weight infants. Acta Paediatr 2010 May;99(5):673–678.

18. Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns. Pediatrics 1994 03;93(3):488–494.

19. Ahlfors CE, Wennberg RP. Bilirubin-albumin binding and neonatal jaundice. Semin Perinatol 2004 10;28(5):334–339.

20. Sinclair CJ. A difference in mortality rate and incidence of kernicterus among premature infants allotted to two prophylactic antibacterial regimens, by William A. Silverman, et al, Pediatrics, 1956;18:614–624. Pediatrics 1998 Jul;102(1 Pt 2):225–227.

9SD

75

Part 2

Recommendations for the management of

hyperbilirubinemia in preterm infants

Chapter 5 Transcutaneous bilirubin measurements in preterm infants: Effects of phototherapy and treatment thresholds 77 Submitted

Chapter 6 Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a national guideline 97 Early Human Development 2011;87:521-5

Chapter 7 High variability and low irradiance of phototherapy devices in Dutch NICUs 113 Arch Dis Child Fetal Neonatal Ed. 2012 May 18. (Epub ahead of print)

Chapter 8 Measurements of neonatal bilirubin and albumin concentrations: A need for improvement and quality control 129 European Journal of Pediatrics. 2011;170:977-82.

67

9SD

5

234

8

77

Chapter 5

Transcutaneous bilirubin measurements

in preterm infants: Effects of photo -

therapy and treatment thresholds

Deirdre E. van Imhoff, Christian V. Hulzebos, Arend F. Bos, Peter H. Dijk

Submitted

78

Abstract

Background: Transcutaneous bilirubin (TcB) levels are increasingly being used in the management of neonatal hyperbilirubinemia. Little is known on the relation between TcB and total serum bilirubin (TSB) in preterm infants receiving photo-therapy (PT). It is also unclear what TcB cut-off level – indicating the need for TSB determination – should be applied in preterm infants.

Aim: To analyze the relation between TSB and TcB levels in preterm infants receiving phototherapy and to determine appropriate TcB cut-off levels.

Methods: TSB and TcB levels were measured daily the first 10 postnatal days in infants of less than 32 weeks of gestation. TcB was measured (Minolta Airshield JM-103) on the covered hipbone within one hour of blood sampling. Agreement between TSB and TcB levels and the mean differences (TSB – TcB) before, during and after PT were assessed. The number of blood samples that could have been reduced and the corresponding false negative measurements were calculated using different methods of TcB cut-off levels.

Results: We obtained 856 paired TcB and TSB levels in 109 preterm infants (66 males, mean (±SD) gestational age 29.4±1.6 weeks and birth weight 1282±316 grams). The mean (±SD) difference between TSB and TcB before PT (44±36 µmol/L) was sig-nificantly lower than during (61±29 µmol/L) and after (63±25 µmol/L) PT (both p<0.01). The measured TcB level plus 50 µmol/L at 70% of the TSB PT threshold resulted in a reduction of 41% of blood samples with 2% false negatives.

Conclusions: Phototherapy increases the underestimation of TSB by TcB in preterm infants. A substantial reduction in the need for TSB measurements with 2% false negative measurements was achieved when the TcB level plus 50 µmol/L at 70% of the TSB PT threshold was used.

Transcutaneous bilirubin measurements in preterm infants 79

Introduction

Neonatal jaundice due to unconjugated hyperbilirubinemia is very common in preterm infants and potentially neurotoxic. The majority of preterm infants of 32 or less weeks of gestational age (GA) develop hyperbilirubinemia.(1) Current treat-ment thresholds for phototherapy (PT) and exchange transfusion (ET) are based on total serum bilirubin (TSB) levels.(2,3,4) As a logical consequence, frequent blood sampling to measure TSB levels is necessary for appropriate management of hyperbilirubinemia in preterm infants. Considering associated side effects and risks of regular blood sampling, i.e. pain, infection and blood loss, alternative non-invasive methods are preferable. Transcutaneous bilirubin (TcB) measurements provide quickly an estimate of TSB levels based on the spectrophotometric measurement of the yellow color of the skin and subcutaneous tissue.(4) Several studies advocate TcB levels as a screening method for hyperbilirubinemia in preterm infants without PT.(5–11) In general, the bleaching effect on the skin by PT is thought to adversely affect the relation between TcB and TSB levels.(12,13) To overcome this problem, a photo-opaque patch attached at the forehead of the infant should be used to protect the skin from the PT light. Data on TcB levels measured on covered skin during PT show good agreement with TSB levels.(14,15,16)

TcB levels may underestimate TSB levels, especially at higher levels.(6,17–19) Consequently, high TSB levels that require treatment (i.e. start, continue or intensify PT or even perform an ET) might be missed. Various methods have been developed to determine TcB cut-off levels that minimize the risk to miss high TSB levels (i.e., false negative TcB measurements) for term and near term infants.(20,21) For preterm infants receiving PT these methods have not been thoroughly studied yet.

Our first aim was to assess the influence of PT on TcB levels, and – more specific – on the relation between TSB and TcB levels in preterm infants. Secondly, we aimed to determine TcB cut-off levels that would reduce the need for blood sampling with minimal risk to miss high TSB levels.

Materials and methods

Patients

The Medical Ethics Committee of the University Medical Center Groningen approved the study. The study was carried out in a subgroup of patients included in a multicenter randomized controlled trial investigating the Bilirubin/Albumin ratio in the treat-ment of preterm infants with hyperbilirubinemia (BARTrial, ISRCTN74465643). The subgroup consisted of infants included between 2007 and 2008 at the neonatal

80

intensive care unit of the Beatrix Children’s hospital (University Medical Center Groningen). Written informed consent was given by the parents or guardians of all infants. Inclusion criteria were a GA of less than 32 weeks and admittance to a level III Neonatal Intensive Care Unit (NICU) within 24 hours after birth. Exclusion criteria were major congenital malformations, clinical syndromes or chromosomal abnormalities.

Measurement of TSB and TcB Levels

We prospectively determined TSB and TcB levels daily during the first 10 days of life. Additional TSB measurements were performed on request of the attending neonatologist.

Within one hour before or after blood sampling for TSB measurement, the TcB measurement was performed with the Minolta Airshields Jaundice Meter 103 ( JM-103, Dräger Medical, Lübeck, Germany). The JM-103 determines the difference in skin reflectance between optical densities for light in the blue (450 nm) and green (550 nm) wavelength regions. By using two optical paths, the reflectance of melanin, dermal maturity and hemoglobin from the superficial tissue can be deducted. The measured TcB level reflects the (sub)cutaneous bilirubin content corrected for gestational age and ethnicity.(22) All NICU nurses were instructed to measure TcB on the hipbone, under the diaper of the infant. The JM-103 displayed the mean TcB level of 3 con-secutive measurements in µmol/L. Regular calibration of the JM-103 following the instructions of the manufacturer was ensured.

TSB was determined using the modified diazo method (Roche Modular, Roche Diagnostics, Woerden, The Netherlands). Blood tubes were stored in the dark until measurement.

Phototherapy

Phototherapy was indicated according to the Dutch TSB PT thresholds for preterm infants of less than 35 weeks of GA with hyperbilirubinemia.(2) These TSB thresholds are categorized in 5 birth weight groups (< 1000 grams, 1000 – 1249 grams, 1250 – 1499 grams, 1500 – 2000 grams and > 2000 grams). PT was registered in hours of use.

At forehand, we measured the irradiance level under a disposable diaper using a Dale40 Phototherapy Radiometer (Fluke Biomedical, Everett, WA, USA). A Draeger 4000 PT device was placed on top of an incubator at 40 cm distance of the diaper. The irradiance level at the hipbone under the diaper was 0.01 – 0.02 µW/cm2/nm compared to 13 µW/cm2/nm without the diaper.

Transcutaneous bilirubin measurements in preterm infants 81

Statistics

The overall correlation of TSB and TcB levels was calculated using all measurements, and subsequently we calculated the correlation between TSB and TcB levels before, during and after PT. The mean TSB and the mean difference between TSB and TcB levels was calculated and illustrated in a Bland Altman plot for all levels and for levels before, during and after PT.(23) To determine significant differences between groups, ANOVA including post hoc multiple comparison analysis of Bonferroni was used. Throughout the manuscript ‘differences’ and ‘mean differences’ between TSB and TcB levels are used interchangeably.

The mean difference between TSB and TcB levels was calculated for different birth weight categories (<1000, 1000–1249, 1250–1499, 1500–2000 and >2000 grams), for different gestational ages (categorized in <26, 26+1 – 28, 28+1 – 30 and 30+1 – 32 weeks) and for the time in hours after PT was stopped. Significant differences were determined using ANOVA including post hoc multiple comparison analysis of Bon-ferroni. A p-value < 0.05 was considered statistically significant.

Different methods to determine TcB cut-off levels were evaluated. TcB cut-off levels are TcB levels at which ‘invasive’ TSB measurements are needed to determine whether treatment is necessary and are based on the Dutch PT thresholds.(2) We used the TcB measurements after 48 hours to define appropriate TcB cut-off levels, because of the increasing slope of the PT thresholds in the first 24–48 postnatal hours. We evaluated the following methods of TcB cut-off levels per birth weight category according to the methods that have been proposed in literature: 1. TcB cut-off levels equivalent to the TSB PT thresholds using the measured TcB, 2. TcB cut-off levels equivalent to the TSB PT thresholds using the measured TcB plus 50 µmol/L and thus correcting for the reported underestimation of the TSB level by the JM-103.(6,17,18,19) 3. TcB cut-off levels at 70% of the TSB PT thresholds considering the large variation in differences between TcB and TSB.(21)(20) and 4. A combination of method 2 and 3: TCB cut off levels equivalent to measured TcB levels plus 50 µmol/L at 70% of TSB PT thresholds.

The Sensitivity, the number and percentage of blood samples that would have been reduced as well as the number and percentage of TcB measurements that missed a high TSB level (false negatives) were calculated. Sensitivity was calculated from the true positive TcB measurements divided by the true positives plus the false negatives multiplied by 100%. The numbers of blood samples that would have been reduced is represented by the true negative (TN) TcB measurements (i.e the number of TcB and TSB measurements lower than the TSB PT threshold). False negative (FN) TcB measurements are TcB levels that inadvertently missed high TSB levels which require treatment (i.e. start, continue or intensify PT or even an ET). False negative

82

TcB measurements were calculated as follows: the number of TcB levels that were below the TSB PT threshold while the actual TSB level was above this threshold. The percentage FN was calculated as percentage of all negative measurements (= (FN/(TN+FN))*100) .

Microsoft Office Excel (Microsoft corporation, Redmond, Washington) and SPSS for Windows (version 18.0 and 20.0, Chicago, IL) were used for data entry and analysis. 

Results

Measurements of 109 infants were included in the analysis (66 males, 88% Cauca-sian). The mean (±SD) gestational age was 29.4 ± 1.6 weeks and the mean (±SD) birth weight was 1282 ± 316 grams. A total of 98 (90%) infants received PT with a mean (±SD) duration of 68 (±52) hours. Eleven infants did not receive PT during the study period.

A total of 856 paired TSB and TcB levels were obtained. Table 1 shows the TSB, TcB and mean differences between TSB and TcB levels for all measurements and for measurements before, during and after PT as well as these levels per birth weight and GA category. TcB levels were significantly higher before PT compared to during and after PT. With increasing birth weight TSB and TcB levels increased significantly (p< 0.01). In addition, TSB and TcB levels were significantly higher in infants of 30+1 – 32 weeks of GA. Figure 1 shows the postnatal course of the TSB and TcB levels in all preterm infants. TSB and TcB show a more or less similar course, though TcB levels are consistently lower than TSB levels, with peak TSB levels reached at day 3 and peak TcB levels reached at day 2 and 3. Figure 2 shows the agreement and the correlation between TSB and TcB levels for all 856 measurements. Almost all (96%) TcB levels except 35 (4%) underestimated the TSB level with a mean (± SD) of 57 ± 31 µmol/L. We found statistically significant and similar correlations between TSB and TcB before, during and after PT (R = 0.81, 0.81 and 0.84, respectively). Figure 3 shows the agreement between TSB and TcB before, during and after PT. The difference between TSB and TcB was significantly lower before PT (44 ± 36) than during and after PT (61 ± 29 µmol/L and 63 ± 25 µmol/L, respectively; p<0.01; Table 1 and Figure 3).

Transcutaneous bilirubin measurements in preterm infants 83

Table 1. Total serum bilirubin and transcutaneous bilirubin levels

TSB (µmol/L) TcB (µmol/L) Mean difference TSB – TcB (µmol/L)

Total (n=865) 134±48 77±51 57±31

Phototherapy

Before PT (n=229) 139±61 95±56 # 44±36*

During PT (n=335) 133±42 72±50 61±29

After PT (n=292) 131±43 68±44 63±25

Birth weight

<1000 (n=149) 97±34 ‡ 42±43 Ω 54 ±28

1000–1249 (n=333) 120±39 ‡ 63±43 Ω 57±31

1250–1499 (n=192) 147±40 ‡ 93±45 Ω 54±30

1500–2000 (n=164) 171±38 ‡ 110±44 Ω 62±33

>2000 (n=18) 228±58 ‡ 152±64 Ω 76±41 †

Gestational age

<26 (n=11) 108±34 45 ±41 63±39

26+1 – 28 (n=204) 118±37 59±46 59±32

28+1 – 30 (n=311) 119±47 67±48 52±30 °

30+1 – 32 (n=330) 158±45 @ 98±50$ 61±30

PT = phototherapy, n = number. TSB = Total Serum Bilirubin, TcB = Transcutaneous Bilirubin, 17.1 µmol/L = 1 mg/dL bilirubin. Data are represented as mean±SD.

Phototherapy:# p<0.01, before PT versus during and after PT

* P<0.01, before PT versus during and after PT

Birth weight:‡ p<0.01, TSB levels per birth weight category Ω p<0.01, TcB levels per birth weight category † p<0.01, infants with a birth weight of > 2000 grams versus 1250 to 1499 grams

Gestational age:@ p<0.01 for infants of 30–32 weeks versus all other GA categories $ p<0.01, for infants of 30–32 weeks versus all other GA categories ° p<0.05, infants with a GA of 28+1 – 30 weeks versus 26+1 – 28 weeks of GA and

p<0.01 versus 30+1 – 32 weeks of GA

84

0

0

100

200

300

1

Postnatal days

TcB µmol/L

TSB µmol/L

TS

B and

TcB

(µm

ol/

L)

2 3 4 5 6 7 8 9 10 11

*

*

Figure 1. The postnatal course of Total Serum Bilirubin and Transcutaneous Bilirubin TSB = Total Serum Bilirubin, TcB = Transcutaneous bilirubin in µmol/L (17.1 μmol/L=1 mg/dL bilirubin). The median is marked by the horizontal line in the central box. The boxes are limited by the 25th and 75th percentiles. The whiskers (┴) represent the lowest and highest TSB and TcB levels within 1.5 interquartile distance below or above the box. Outliers (○) represent TSB and TcB levels between 1.5 and 3 interquartile distances below or above the box. Extremes (*) represent TSB and TcB levels more than 3 interquartile distances below or above the box.

Transcutaneous bilirubin measurements in preterm infants 85

Figu

re 2.

Bla

nd A

ltman

and

corr

elat

ion

plot

for a

ll 85

6 le

vels

A.

The a

gree

men

t bet

wee

n T

SB an

d Tc

B de

pict

ed in

a Bl

and-

Altm

an p

lot.

The X

-axi

s sho

ws t

he m

ean

of T

SB an

d Tc

B in

µm

ol/L

and

the Y

-axi

s sh

ows t

he d

iffer

ence

bet

wee

n T

SB an

d Tc

B in

µm

ol/L

. The h

oriz

onta

l lin

e rep

rese

nts t

he m

ean

diffe

renc

e (57

µm

ol/L

), th

e dott

ed li

nes

repr

esen

t the

95%

confi

denc

e int

erva

l (- 5

to 11

9 µm

ol/L

) B.

Th

e cor

rela

tion

betw

een

TSB

and

TcB

in µ

mol

/L.

17.1µ

mol

/L =

1 m

g/dL

bili

rubi

n

0

100 0

200

30

0

40

0

100

TS

B ( µ

mol/

L)

TcB (µmol/L)

200

40

030

0

B

R2 =

0.6

5

0

0

100

100

200

30

0

100

Mean T

SB

and

TcB

(µm

ol/

L)

Difference TSB-TcB (µmol/L)

200

30

0

A

86

0

0

100

100

200

100

Mean T

SB

and

TcB

(µm

ol/

L)

Difference TSB-TcB (µmol/L)

200

30

0

Befo

re P

T

0

0

100

100

200

100

Mean T

SB

and

TcB

(µm

ol/

L)

Difference TSB-TcB (µmol/L)

200

30

0

Duri

ng

PT

Transcutaneous bilirubin measurements in preterm infants 87

0

0

100

100

200

100

Mean T

SB

and

TcB

(µm

ol/

L)

Difference TSB-TcB (µmol/L)

200

30

0

Aft

er

PT

Figu

re 3.

Bla

nd A

ltman

plo

ts b

efor

e, du

ring,

and

after

pho

toth

erap

y Th

e agr

eem

ent b

etw

een

TSB

and

TcB

depi

cted

in a

Blan

d-A

ltman

plo

t bef

ore,

durin

g and

after

PT.

The X

-axi

s sho

ws t

he m

ean

of T

SB an

d Tc

B in

µm

ol/L

and

the Y

-axi

s sho

ws t

he d

iffer

ence

bet

wee

n T

SB an

d Tc

B in

µm

ol/L

. The h

oriz

onta

l lin

e rep

rese

nts t

he m

ean

diffe

renc

e; th

e dott

ed li

nes

repr

esen

t the

95%

confi

denc

e int

erva

l. 17

.1µm

ol/L

= 1

mg/

dL b

iliru

bin

88

The difference between TSB and TcB levels was similar for all birth weight categories except for infants weighing more than 2000 grams. The difference of TSB and TcB was significantly higher in this group compared to infants of 1250 to 1499 grams.

It appeared that the difference between TcB and TSB was affected by GA. The difference between TSB and TcB levels for infants with a GA of 28+1–30 weeks was significantly lower (52±30) than the mean difference of infants of 26+1 – 28 and 30+1 – 32 weeks of GA (59±32 µmol/L and 61±30 µmol/L, p< 0.05 and p<0.01, respectively; Table 1). The difference between TSB and TcB was statistically not significant affected by the time after PT was discontinued. After 6 h (n=5), 12 h(n=54), 24 h(n=47) and 48 h(n=101) or more hours after PT was stopped, the differences between TSB and TcB levels were 46 ± 23, 60 ± 22, 66 ± 19 and 64 ± 20 µmol/L, respectively.

Table 2 shows the sensitivity, the true negatives (blood samples that would have been reduced), and false negatives (TcB measurements that missed a high TSB level) for the different birth weight categories and for the different TcB cut-off levels. When TcB cut-off levels equivalent to PT thresholds were used, the mean percentage of blood samples in all birth weight categories that would have been reduced was 80%. However, the corresponding overall false negative rate was as high as 16%. To correct for the underestimation 50 µmol/L was added to the measured TcB levels (‘TcB + 50 µmol/L’). This resulted in a lower percentage number of blood samples that would have been reduced: 73% with 9% false negatives. Considering the broad variation of differences between TSB and TcB levels, TcB cut-off levels at 70% of the TSB PT thresholds were used. This would have reduced the need for blood sampling with 73% with a corresponding false negative rate of 9%. The combination of ‘TcB + 50 µmol/L’ at 70% of the TSB PT threshold resulted in a reduction for the need of blood sampling of 41% (Table 2) and reduced the false negatives to 2%. Sensitivity was highest (88 to 100%) when a combination of TcB + 50 µmol/L at 70% of the TSB PT threshold was used.

Transcutaneous bilirubin measurements in preterm infants 89

Tabl

e 2 A

and

B. E

ffect

of d

iffer

ent T

cB cu

t-off

leve

ls pe

r birt

h w

eigh

t cat

egor

y on

the s

ensit

ivity

, blo

od sa

mpl

es th

at w

ould

hav

e bee

n re

duce

d an

d on

th

e cor

resp

ondi

ng T

cB m

easu

rem

ents

miss

ing a

hig

h T

SB le

vel

A Ap

plie

d T

cB c

ut-

off

leve

lTc

BTc

B + 5

0

Birt

h w

eig

ht

(g)

NT

SB

thre

shold

Sensi

tivi

tyRed

uct

ion

T

NM

issi

ng

FN

Sensi

tivi

tyRed

uct

ion

T

NM

issi

ng

FN

≤ 9

99

125

100

25%

88(7

0%

)27

(23%

)61%

80

(64

%)

14(1

5%

)

100

0 –

124

929

715

06

%22

9(7

7%)

62(

21%

)50

%20

7(70

%)

33(1

4%

)

1250

– 14

49

168

190

9%

145(8

6%

)21

(13%

)61%

137(

82%

)9

(6%

)

150

0 –

19

99

134

220

82%

117(

87%

)3(3

%)

41%

109

(81%

)10

(8%

)

≥ 20

00

16

240

25%

8(5

0%

)6

(43%

)6

3%

8(5

0%

)3(2

7%)

All

(mean)

740

80

%16

%73

%9

%

B Ap

plie

d T

cB c

ut-

off

leve

lTc

BTc

B + 5

0

Birt

h w

eig

ht

(g)

N70

% T

SB

thre

shold

Sensi

tivi

tyRed

uct

ion

T

NM

issi

ng

FN

Sensi

tivi

tyRed

uct

ion

T

NM

issi

ng

FN

≤ 9

99

125

70

50

%85(6

8%

)18

(17%

)89

%47(

38

%)

4(8

%)

100

0 –

124

929

710

54

5%

210

(71%

)36

(15%

)89

%12

9(4

3%

)7(

5%

)

1250

– 14

49

168

133

70%

131(

78%

)7(

5%

)10

0%

70(4

2%)

0

150

0 –

19

99

134

154

59

%10

7(80

%)

7(6

%)

88%

53(4

0%

)2(

4%

)

≥ 20

00

16

168

75%

6(3

8%

)2(

25%

)88%

5(3

1%)

1(17

%)

All

(mean)

740

73%

9%

41%

2% N

= n

um

ber

of

measu

rem

ents

in t

he c

orr

esp

ond

ing

bir

th w

eig

ht

cate

gory

. TS

B = T

ota

l Ser

um

Bili

rub

in (

17.1

µmol/

L = 1 m

g/d

L b

iliru

bin

), T

cB =

Tra

nsc

uta

neous

bili

rub

in, B

irth

weig

ht

in g

ram

s.

True p

osi

tive

(T

P), t

rue n

eg

ati

ve (

TN

), fa

lse p

osi

tive

(FP)

and

fa

lse n

eg

ati

ve (

FN

) m

easu

rem

ents

wer

e u

sed

to c

alc

ulate

the s

ensi

tivi

ty, t

he r

ed

uct

ion in

blo

od

sa

mp

les

and

the T

cB m

easu

rem

ents

th

at

mis

sed

a T

SB

leve

l th

at

woul

d h

ave

ind

icate

d a

ch

ang

e in

ther

ap

y (s

tart

, conti

nue o

r in

tensi

fy P

T, E

T)

(mis

sing

).

Sensi

tivi

ty (

%)

= T

P/(T

P+FN

) *

100

, Red

uct

ion =

TN

, red

uct

ion %

= T

N/A

LL *

10

0, M

issi

ng

= F

N, m

issi

ng

% =

FN

/(FN

+T

N)

* 10

0

A. T

SB t

hre

shold

s re

flect

the T

SB le

vel o

f PT

aft

er 4

8 p

ost

nata

l hours

.

B. 7

0%

TS

B th

resh

old

s re

flect

70

% o

f T

SB

leve

l of

PT a

fter

48 p

ost

nata

l hours

.

90

Discussion

In the present study we found that TcB levels measured with the JM-103 show a strong correlation and rather good agreement with TSB levels in preterm infants with and without PT. The data also show that TcB consistently underestimates the TSB level with a mean of 57 µmol/L and that phototherapy does affect this underestimation of TSB: the underestimation increases, and this effect remains after PT is stopped. Finally, we present results of a novel TcB cut-off method which corrects for this un-derestimation and for the variation in differences between TSB and TcB levels, with minimal risks to miss infants with significant hyperbilirubinemia.

Our data confirm the previously reported underestimation of TSB concentrations by TcB levels measured with the JM-103 device. Ebbesen et al. found a median differ-ence between TSB and TcB of 67 µmol/L in preterm infants (GA 24 – 34+6 weeks) without PT.(24) Moreover, the NICE (National Institute for health and Clinical Excellence) ‘neonatal jaundice’ guideline group expressed their concern about this underestimation of 50 µmol/L with this particular device.(6)(17,18,19)

Limited data exist on the effect of PT on the accuracy of TcB measurements us-ing the JM-103 device in preterm infants. Zecca et al. studied TcB measurements in preterm infants using the BiliCheck combined with a photo-opaque skin patch glued to the forehead of the infant. They found that the mean difference between TSB and TcB measurements on the covered skin was only 3.4 µmol/L compared to 55 µmol/L (p<0.001) for TcB measurements on the exposed skin.(14) Nanjundaswamy et al. used the BiliCheck device combined with a standard PT eye patch on the forehead of preterm infants. They confirmed that PT light was not transmitted through the eye patch. The correlation between TSB and TcB levels was weaker for TcB measure-ments on the exposed skin compared to measurements on the covered skin (R = 0.70 versus R = 0.77, respectively; p<0.04) and compared to the control measurements before PT (R = 0.70 versus R = 0.86, respectively; p<0.01).(15) Likewise, Knupfer et al. described a significant weaker correlation between TSB and TcB levels for preterm infants with PT compared to that for infants without PT (R = 0.59 versus R = 0.73, respectively; p<0.05). Measurements were performed with the BiliCheck, but they did not mention whether TcB measurements were performed on the covered or ex-posed skin.(25) Studies on the effect of PT on TcB measurements using the JM-103 are confined to term infants. Tan et al. found that TcB measurements in term infants using the JM-103 during PT on the covered skin showed a good, but weaker correla-tion with TSB levels (R = 0.74) compared to TcB measurements in controls without PT (R = 0.80). In contrast to the previously described studies in preterm infants,

Transcutaneous bilirubin measurements in preterm infants 91

Tan et al. analyzed TSB and TcB levels also after cessation of PT: The effect of PT on the TcB measurements disappeared 18 to 24 hours after PT was stopped.(13,14,15,25)

In the present study we found similar TcB -TSB correlations before, during and after PT (R = 0.81, 0.81 and 0.84). However, the underestimation of TSB increases significantly after PT is started and remains higher after cessation of PT. This shows that high correlations do not necessarily correspond with sufficient accuracy of TcB measurements, To the best of our knowledge, this study is the first which includes an analysis of agreement. In our opinion, the statistically significant increase in the underestimation during and after PT is relatively small and seems clinically of mod-est relevance.

We measured TcB levels under the diaper on the hipbone of the preterm infants to minimize the bleaching effect of PT on the skin, and to minimize disturbance of the vulnerable preterm skin by glued patches. Although in the present study the ir-radiance level under the diaper was considered minimal (0.01–0.02 µW/cm2/nm), we can not completely exclude some transmittance of PT irradiance, e.g. when the infant moves and the diaper slips a bit.

The present study shows that the mean difference between TSB and TcB levels of infants with a birth weight of more than 2000 grams was significantly higher compared to infants of 1250 to 1499 grams. As a consequence, the false negative TcB rate was consistently the highest for infants with a birth weight of more than 2000 grams, for all TcB cut-off levels that we evaluated (Table 2). It is possible that the small num-ber of measurements in this group (n = 18) resulted in these divergent results. Yet, these results may also be affected by the height of the TSB levels. TcB levels tend to underestimate TSB levels more at higher TSB levels.(18,24) In agreement with this, mean TSB levels in this group of infants were significantly higher compared to other birth weight categories.

The significant lower mean difference in infants of 28+1 to 30 weeks of GA com-pared to infants of 26+1 – 28 and 30+1 – 32 weeks of GA could not be explained by higher TSB levels nor by differences in birth weight and remains sofar unexplicable

Are TcB measurement beneficial for the care of preterm infants with hyperbilirubi-nemia? Data on the number of blood samples that could theoretically be reduced by using TcB levels in preterm infants range from 2–77% with false negative rates between 0 and 56% (Table 3).(9–11,24,25) However, any comparison between these studies is hazardous. Many factors influence TcB measurements: gestational age, postnatal age, ethnicity, measurement site, type of TcB device and its algorithm, the laboratory techniques for TSB measurements, as well as the applied treatment thresholds. For preterm infants with imminent bilirubin neurotoxicity it is obvious to apply TcB cut off levels that minimize the risk of missing high TSB levels. We determined effects

92

Tabl

e 3. S

tudi

es o

n Tc

B cu

t-off

leve

ls an

d pe

rcen

tage

s of b

lood

sam

ples

that

coul

d be

redu

ced

in p

rete

rm in

fant

s with

out P

T

Auth

or

GA

TcB

Dev

ice

TcB

cut-

off

le

vel

Sensi

tivi

tyRed

uct

ion

T

NM

issi

ng

FN

Knup

fer,

200

1(25

)33–3

7 w

eeks

BCAt

the T

SB

PT t

hre

shold

88%

107(

74%

)3(3

%)

Ebb

ese

n, 2

00

2(20

)25

–37

weeks

BCAt

70%

of

the T

SB

PT t

hre

shold

99

%#

13–4

2%1(

5%

)

Will

em

s, 2

00

4 (

10)

26–2

9 w

eeks

BCAt

70%

of

the T

SB

thre

shold

N/A

35(3

7%)

1(3%

)

Karo

lyi,

200

4(1

1)

23–3

3 w

eeks

JM-1

02

At

‘14’*

100

%32(

26%

)0

Sanp

ava

t, 2

00

7(9

)30

–35 w

eeks

JM-1

03

At

the T

SB

PT t

hre

shold

53–9

8%

193–4

(77–

2%)

15–5

(7–5

6%

)

Ebb

ese

n, 2

012

(24

)< 3

5 w

eeks

JM-1

03

At

35%

of

the T

SB

PT t

hre

shold

97%

58(2

4%

)2(

3%

)

GA

= G

est

ati

ona

l ag

e, T

cB =

Tra

nsc

uta

neous

Bilir

ub

in. B

C =

Bili

Check

, JM

= M

inolta

Air

shie

lds.

PT

= p

hoto

ther

ap

y, 17.

1 µm

ol/

L = 1 m

g/d

L b

iliru

bin

. PT

thre

shold

are

TS

B-b

ase

d, T

SB =

Tota

l Ser

um

Bili

rub

in.

Sensi

tivi

ty (

%)

= T

P/(T

P+FN

) *

100

, Red

uct

ion =

TN

, red

uct

ion %

= T

N/A

LL *

10

0, M

issi

ng

= F

N, m

issi

ng

% =

FN

/(FN

+T

N)

* 10

0

N/A

= n

ot

ap

plic

ab

le, i

.e. n

ot

menti

oned

in t

he a

rtic

le o

r not

poss

ible

to c

alc

ulate

fro

m t

he d

ata

giv

en in

the s

tud

y.

# T

he s

ensi

tivi

ty w

as

det

erm

ined

for

all

NIC

U in

fants

incl

ud

ing

infa

nts

of

more

th

an 3

7 w

eeks

of

gest

ati

ona

l ag

e*

Fro

m t

his

stu

dy

the T

cB c

ut-

off

leve

l was

not

oth

erw

ise s

peci

fied

, 14

ma

y re

fer

to ‘r

eflect

ance

unit

s’

Transcutaneous bilirubin measurements in preterm infants 93

of different internationally used methods to determine TcB cut-off levels that aim to reduce the need for blood sampling without missing high TSB levels. We cor-rected for the underestimation by adding 50 µmol/L to the measured TcB level and noticed a marked reduction in the need for blood sampling, but a concomitant high risk to miss significant hyperbilirubinemia. Secondly, we corrected for the variance in differences between TSB and TcB levels by lowering the PT levels to 70% of the original PT threshold. This novel cut-off method resulted in a 41% reduction of blood samples and a false negative rate of 2%. (Table 2)

Based on our data and previously published data we propose recommendations for the use of the transcutaneous bilirubin measurements in preterm infants (Table 4). These recommendations include teaching, applying and controlling the use of TcB measurement in daily clinical practice (Table 4).(6,17,26)

Table 4. Recommendations for the use of JM-103 TcB measurements in preterm infants with and without PT

Recommendations

• Measure TcB levels under the diaper on the hipbone of the infant

• TcB cut-off level: Add 50 µmol/L to the measured TcB level at 70% of the PT threshold

• Ensure regular calibration of the TcB device according to the recommendations of the manufacturer

• Regularly perform paired TSB and TcB levels to check the accuracy of the TcB device

• Teach NICU nurses and attending physicians how to perform and interpret TcB mea-surements

• Measure the TSB level: –If the TcB level plus 50 µmol/L exceeds 70% of the TSB PT threshold –If there is clinical concern on severe hyperbilirubinemia –If you don’t trust the result of that transcutaneous measurement

TSB = total serum bilirubin, TcB = transcutaneous bilirubin, PT = phototherapy, NICU = neonatal intensive care unit, 17.1 µmol/L = 1 mg/dL bilirubin

94

Conclusion

PT increases the underestimation of TSB by TcB, but TcB levels show more or less similar agreement with TSB levels when measured with the JM-103 on the covered hipbone in preterm infants of 32 or less weeks of GA before, during and after PT. In preterm infants we recommend to use a TcB+50 µmol/L cut-off level at 70% of the TSB PT threshold, which is associated with a substantial reduction in the need for blood samples and a minimal risk to miss high TSB levels.

Transcutaneous bilirubin measurements in preterm infants 95

References

1. Maisels MJ. Neonatal hyperbilirubinemia and kernicterus – Not gone but sometimes forgotten. Early Hum Dev 2009 10/14.

2. van Imhoff DE, Dijk PH, Hulzebos CV, on behalf of the BARTrial studygroup of the Netherlands Neonatal Research Network. Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a national guideline. Early Hum Dev 2011 Aug;87(8):521–525.

3. American Academy oP. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004 07;114(1):297–316.

4. Yamanouchi I, Yamauchi Y, Igarashi I. Transcutaneous bilirubinometry: preliminary studies of noninvasive transcutaneous bilirubin meter in the Okayama National Hospital. Pediatrics 1980 02;65(2):195–202.

5. Badiee Z, Mohammadizadeh M, Shamee M. Diagnostic usefulness of transcutaneous bilirubinometry in very preterm newborns. Int J Prev Med 2012 Apr;3(4):262–265.

6. Schmidt ET, Wheeler CA, Jackson GL, Engle WD. Evaluation of transcutaneous bilirubinometry in preterm neonates. J Perinatol 2009 Aug;29(8):564–569.

7. Stillova L, Matasova K, Zibolen M, Stilla J, Kolarovszka H. Transcutaneous bilirubinometry in preterm neonates. Indian Pediatr 2009 May;46(5):405–408.

8. Namba F, Kitajima H. Utility of a new transcutaneous jaundice device with two optical paths in premature infants. Pediatr Int 2007 Aug;49(4):497–501.

9. Sanpavat S, Nuchprayoon I. Transcutaneous bilirubin in the pre-term infants. J Med Assoc Thai 2007 Sep;90(9):1803–1808.

10. Willems WA, van den Berg LM, de Wit H, Molendijk A. Transcutaneous bilirubinometry with the Bilicheck in very premature newborns. J Matern Fetal Neonatal Med 2004 Oct;16(4):209–214.

11. Karolyi L, Pohlandt F, Muche R, Franz AR, Mihatsch WA. Transcutaneous bilirubinometry in very low birthweight infants. Acta Paediatr 2004 Jul;93(7):941–944.

12. Ozkan H, Oren H, Duman N, Duman M. Dermal bilirubin kinetics during phototherapy in term neonates. Acta Paediatr 2003 May;92(5):577–581.

13. Tan KL, Dong F. Transcutaneous bilirubinometry during and after phototherapy. Acta Paediatr 2003;92(3):327–331.

14. Zecca E, Barone G, De Luca D, Marra R, Tiberi E, Romagnoli C. Skin bilirubin measurement during phototherapy in preterm and term newborn infants. Early Hum Dev 2009 08;85(8):537–540.

15. Nanjundaswamy S, Petrova A, Mehta R, Hegyi T. Transcutaneous bilirubinometry in preterm infants receiving phototherapy. Am J Perinatol 2005 Apr;22(3):127–131.

16. Jangaard K, Curtis H, Goldbloom R. Estimation of bilirubin using BiliChektrade mark, a transcutaneous bilirubin measurement device: Effects of gestational age and use of phototherapy. Paediatr Child Health 2006 Feb;11(2):79–83.

96

17. National Institute for Health and Clinical Excellence. National Institute for Health and Clinical Excellence. Neonatal Jaundice (Clinical guideline 98). www.nice.org.uk/CG98. 2010.

18. Engle WD, Jackson GL, Stehel EK, Sendelbach DM, Manning MD. Evaluation of a transcutaneous jaundice meter following hospital discharge in term and near-term neonates. J Perinatol 2005 Jul;25(7):486–490.

19. Chang Y, Hsieh W, Chou H. The effectiveness of a noninvasive transcutaneous bilirubin meter in reducing the need for blood sampling in Taiwanese neonates. clinical neonatology 2006;13(2):60–63.

20. Ebbesen F, Rasmussen LM, Wimberley PD. A new transcutaneous bilirubinometer, BiliCheck, used in the neonatal intensive care unit and the maternity ward. Acta Paediatr 2002;91(2):203–211.

21. Maisels MJ, Bhutani VK, Bogen D, Newman TB, Stark AR, Watchko JF. Hyperbilirubinemia in the newborn infant > or =35 weeks’ gestation: an update with clarifications. Pediatrics 2009 10;124(4):1193–1198.

22. Yasuda S, Itoh S, Isobe K, Yonetani M, Nakamura H, Nakamura M, et al. New transcutaneous jaundice device with two optical paths. J Perinat Med 2003;31(1):81–88.

23. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986 Feb 8;1(8476):307–310.

24. Ebbesen F, Vandborg PK, Trydal T. Comparison of the transcutaneous bilirubinometers BiliCheck and Minolta JM-103 in preterm neonates. Acta Paediatr 2012 Nov;101(11):1128–1133.

25. Knupfer M, Pulzer F, Braun L, Heilmann A, Robel-Tillig E, Vogtmann C. Transcutaneous bilirubinometry in preterm infants. Acta Paediatr 2001 Aug;90(8):899–903.

26. Maisels MJ, Ostrea EM,Jr., Touch S, Clune SE, Cepeda E, Kring E, et al. Evaluation of a new transcutaneous bilirubinometer. Pediatrics 2004 06;113(6):1628–1635.

5

7

9SD

6

234

8

97

Chapter 6

Uniform treatment thresholds for

hyperbilirubinemia in preterm infants:

Background and synopsis of a national

guideline

Deirdre E. van Imhoff, Peter H. Dijk, Christian V. Hulzebos

and on behalf of the BARTrial studygroup of the Netherlands Neonatal Research network

Early Human Development 2011;87:521–5

98

Abstract

Background: To prevent severe hyperbilirubinemia and bilirubin neurotoxicity, the American Academy of Pediatrics’ management guideline for hyperbilirubinemia in near term infants is used world wide. A leading guideline for jaundiced preterm in-fants is lacking whereas the risk on severe hyperbilirubinemia is high in these infants. Our aim was to define uniform treatment thresholds for jaundiced preterm infants. In this article we present the history and a synopsis of this novel national guideline.

Study design: A survey on guidelines for hyperbilirubinemia in preterm infants was sent to all Dutch Neonatal Intensive Care Units (NICUs). After comparison with international guidelines, a new consensus-based guideline was developed.

Results: Treatment thresholds of all 10 NICUs were based on Total Serum Bilirubin (TSB) and related to birth weight (n=9) and gestational age (n=1). NICUs used age-specific (n=6) or fixed (n=4) TSB-thresholds resulting in a large range of thresholds (maximal 170 µmol/L for phototherapy and 125 µmol/L for exchange transfusion). Acidosis, asphyxia, sepsis, active hemolysis and intraventricular hemorrhage were most frequently used risk factors. Consensus was agreed upon TSB-based treatment thresholds, categorized in 5 birth weight groups and divided in high and low risk infants.

Conclusion: There was no standardized care for jaundiced preterm infants in the Netherlands. In addition to the internationally used guideline for (near) term in-fants, a novel ‘consensus based’ guideline for preterm infants with a gestational age of less than 35 weeks has been developed and implemented in The Netherlands. This guideline is approved and recommended by the Dutch Society of Pediatrics.

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 99

Introduction

Neonatal jaundice due to unconjugated hyperbilirubinemia is very common in term and preterm infants. Despite effective treatment strategies for hyperbilirubinemia, i.e. phototherapy and exchange transfusion, bilirubin induced neurotoxicity still occurs.(1–4) 

To reduce the incidence of severe neonatal hyperbilirubinemia and bilirubin-induced neurological damage, the American Academy of Pediatrics’ (AAP) recommended a guideline for the prevention and management of unconjugated hyperbilirubinemia in infants of 35 or more weeks of gestational age (GA). This AAP-hyperbilirubinemia guideline has been adopted by many hospitals world-wide including the Netherlands.(3,5,6) Uniform guidelines for hyperbilirubinemia in preterm infants (of 35 or less weeks of GA) are lacking. The limited evidence on safe treatment thresholds and risk factors for preterm infants is reflected in the well-known large variation in applied (inter)national guidelines.(1,3,7–16)

Preterm infants are more at risk for developing bilirubin encephalopathy compared to their term counterparts. Kernicterus, even in the absence of classical neurological signs, has been described in preterm infants at total serum bilirubin (TSB) con-centrations which were generally considered safe.(1) Therefore, lower TSB- based treatment thresholds are usually applied in preterm infants. Yet, legitimate evidence on these TSB-thresholds is lacking and many different guidelines are used.(10,17,18) Our aim was to define uniform TSB-based treatment thresholds and risk factors for hyperbilirubinemia in preterm infants.

Methods

In 2006, a survey was sent to all 10 Dutch Neonatal Intensive Care Units (NICUs) on their local guidelines for the management of hyperbilirubinemia in preterm infants. Treatment thresholds for phototherapy and exchange transfusion were analyzed as well as criteria determining the height and slope of the TSB-threshold curve i.e. birth weight, postnatal age and applied risk factors. TSB treatment thresholds were categorized in 3 birth weight groups, postnatal age in hours and 2 (high or standard) risk groups for analysis in Microsoft Office Excel (Microsoft corporation, Redmond, Washington) and SPSS for Windows (version 16.0, Chicago, IL).

After comparison with (inter)national guidelines on hyperbilirubinemia, a novel consensus-based guideline was developed.

100

Results

All NICUs (n=10) responded to our request. Treatment thresholds were based on TSB in all 10 NICUs. One NICU used gestational age and 9 NICUs used birth weight to determine TSB-thresholds. Birth weight was categorized in 2 (n=1), 3 (n=7) or 5 (n=1) groups. Postnatal age in hours determined TSB-thresholds in 6 NICUs. The remaining 4 NICUs used fixed TSB-thresholds. After day 4, fixed TSB-thresholds were used in all NICUs.

Risk factors were incorporated in all NICU guidelines, although large variability in the applied risk factors existed. Table 1 shows the fourteen risk factors used in the NICUs. Sepsis, acidosis (pH <7.20), asphyxia (Apgar score < 6 at five minutes or umbilical cord pH < 7.10), active hemolysis and intraventricular hemorrhage were most frequently used.

Table 1. Risk factors used in the Netherlands in the management of unconjugated hyperbilirubinemia in preterm infants

Risk factors Number of NICUs

Acidosis (pH < 7.25) 9

Sepsis 8

Asphyxia 8

Active hemolysis 7

Intraventricular hemorrhage 5

Cerebral ultrasonography corresponding with asphyxia 3

Albumin < 20 g/L 3

Hypoxemia 2

Fluctuating temperature 2

IRDS 1

Convulsions 1

Lethargy 1

Fluid restriction 1

Rapid rise in TSB (> 17 µmol/L/hour) 1

Asphyxia was defined as Apgar < 6 after 5 minutes or umbilical cord pH < 7.10, hypoxemia was de-fined as an arterial pO2 < 5.0 kPa for more than 2 hours. NICU = Neonatal Intensive Care Unit, IRDS = Infant Respiratory Distress Syndrome, TSB = Total Se-rum Bilirubin in µmol/L (17.1 µmol/L = 1 mg/dL).

Table 2 shows the ranges in TSB-thresholds for phototherapy and exchange transfu-sion used in the 10 Dutch NICUs; TSB-thresholds of the first 4 postnatal days were categorized in the 3 most commonly used birth weight groups, and in standard or high risk infants. Maximum TSB ranges were consequently found on day 4 (T = 96 hours)

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 101

Tabl

e 2. R

ange

s in

TSB

-thre

shol

ds in

the N

ethe

rland

s for

pre

term

infa

nts w

ith u

ncon

juga

ted

hype

rbili

rubi

nem

ia

Birt

h W

eig

ht

Ris

kT

=24

hT

=48h

T=7

2hT

=96

h

PTET

PTET

PTET

PTET

<10

00

Sta

ndard

100

–155

17

0–2

2010

0–1

85

170

–255

100

–20

517

0–2

80

100

–220

170

–29

0

Hig

h70

–120

14

0–2

00

85–1

55

170

–230

85–1

7517

0–2

50

85–1

85

170

–255

100

0–1

50

0

Sta

ndard

100

–185

17

0–2

7510

0–2

2021

5–2

7010

0–2

35

215–3

05

100

–255

215–3

25

Hig

h85–1

55

17

0–2

2085–1

85

175–2

55

85–2

05

175–2

80

85–2

2017

5–2

90

150

0–2

50

0

Sta

ndard

130

–220

205–3

00

140

–255

255–3

1014

0–2

80

275–3

40

140

–29

027

5–3

55

Hig

h85–1

85

17

0–2

50

85–2

2020

0–2

7085–2

45

200

–30

58

5–2

55

200

–325

Tota

l ser

um

bili

rub

in (

TS

B in

µm

ol/

L) r

ang

es

for

photo

ther

ap

y (P

T)

and

exc

hang

e tra

nsf

usi

on (

ET)

exp

ress

ed

in p

ost

nata

l ag

e (

T)

in h

ours

(h)

and

ca

teg

ori

zed

in s

tand

ard

ris

k (S

tand

ard

) and

hig

h r

isk

(Hig

h).

Bold

va

lues

rep

rese

nt

the h

ighest

rang

es

in T

SB-

thre

shold

s.

102

for phototherapy thresholds. Ranges of 120, 155 and 150 µmol/L were found for preterm infants with a standard risk and a birth weight of < 1000 grams, 1000 – 1500 grams or 1500 – 2500 grams, respectively. For infants considered to have high risk factors and birth weights of <1000 grams, 1000 – 1500 grams or 1500 – 2000 grams, maximum ranges were 100, 135 and 170 µmol/L, respectively.

Figure 1 shows phototherapy and exchange transfusion thresholds for standard risk infants with a birth weight of 1000 to 1500 grams.

0

100

200

300

Age (hrs)

Phototherapy

Exchangetransfusion

TS

B (µ

mol/

L)

24 48 72 96

Figure 1. TSB-thresholds of phototherapy and exchange transfusion in the Netherlands for standard risk preterm infants with a birth weight between 1000 and 1500 grams Ranges in TSB-thresholds used in the Netherlands for phototherapy and exchange transfusion (17.1 µmol/L = 1 mg/dL) for the postnatal age (hours) of the preterm infant. The median is marked by the horizontal line in the central box. The boxes are limited by the 25th and 75th percentiles. The whiskers (┴) represent the lowest and highest TSB-thresholds within 1.5 interquartile distance below or above the box. Outliers (○) represent TSB-thresholds between 1.5 and 3 interquartile distances below or above the box.

Based on these results and on published recommendations for treatment of un-conjugated hyperbilirubinemia in preterm infants, consensus between the Dutch

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 103

NICUs was reached and uniform guidelines were developed.(10,17) Arbitrarily, 5 birth weight groups were defined (i.e. < 1000 g, 1000 – 1250 g, 1250 – 1500 g, 1500 – 2000 g and > 2000 g) as well as most commonly applied risk factors on either the risk of hyperbilirubinemia or the risk on bilirubin neurotoxicity (i.e. active hemolysis (with positive Coombs), asphyxia, hypoxemia, acidosis, and clinical/neurological deterioration (sepsis/meningitis or intracranial hemorrhage > grade 2 according to Papile). All 5 birth weight nomograms are shown in Figure 2 and on https://www.babyzietgeel.nl/index.php?id=135.

Figure 2. Phototherapy and exchange transfusion thresholds for preterm infants of less than 35 weeks of gestation TSB-thresholds of phototherapy (PT) and exchange transfusion (ET) in preterm infants of 35 or less weeks of gestational age. TSB-thresholds (17.1 µmol/L = 1 mg/dL) versus postnatal age (days). Standard or high risk is based on presence of risk factors. All nomograms can be downloaded from http://www.babyzietgeel.nl/index.php?id=135.

104

After a gradual increase of the TSB-threshold in the first 24–48 postnatal hours a plateau is reached. Since 2008, this guideline is approved and recommended by the Dutch Society of Pediatrics to use for all preterm infants of less than 35 weeks of gestational age in the Netherlands.

Discussion

Large variability existed in TSB treatment thresholds for unconjugated hyperbiliru-binemia in preterm infants in the Netherlands. Age-specific TSB-thresholds were not used in 4 NICUs, which partially explains the large range of applied TSB-thresholds. 

Our results are in agreement with previous (inter)national data and support the lack of evidence on specific TSB-thresholds and neurological outcome.(1,7–9,14,15,19–23) To the best of our knowledge, only 3 prospective studies have analyzed effects of different TSB-thresholds on biochemical, phototherapy, and/ or outcome data in preterm infants (Table 3).(24–26) In 1985, Curtis-Cohen et al. randomly assigned 22 preterm infants with a birth weight of less than 1250 gram to a prophylactic or a conservative treatment. Phototherapy was initiated immediately postnatal in the prophylactic group (n = 11), whereas in the conservative group phototherapy was started above TSB concentrations of 85 µmol/L. Mean TSB concentrations at onset of phototherapy were significantly different (102 µmol/L versus 39 µmol/L, conser-vative versus prophylactic phototherapy, respectively; p < 0.001) and total duration of phototherapy was 48 hours longer in the prophylactic group (p<0.05), whereas maximum TSB concentrations, age at maximum TSB concentrations and rate of rise of TSB were similar.(24) Long-term neurodevelopmental outcome has been ana-lyzed in two prospective trials on prophylactic versus conservative phototherapy in preterm infants. Jangaard et al. started prophylactic phototherapy (n = 46) 12 hours after birth, and conservative phototherapy (n = 49) at a predefined TSB concentra-tion of 150 µmol/L in newborn infants with a birth weight lower than 1500 gram. Maximum TSB concentrations did not differ significantly except in a subgroup of infants weighing less than 1000 grams: Maximum TSB concentrations were highest in the conservative group (171 µmol/L versus 139 µmol/L, conservative versus pro-phylactic phototherapy, respectively, p < 0.02). A non-significant tendency towards poor long term neurodevelopmental outcome in conservative-treated extreme low birth weight infants was reported.(25) Recently, a randomized controlled trial in-vestigated the effects of prophylactic phototherapy (started immediately postnatal) versus conservative phototherapy (based on predefined TSB concentrations) on outcome in hyperbilirubinemic preterm infants with a birth weight of < 1000 grams. Maximum TSB concentrations were higher in the conservative group (168 µmol/L

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 105

Tabl

e 3. S

tudi

es o

n T

SB th

resh

olds

and

outc

ome

Auth

or/

Ye

ar

Pop

ulati

on

TS

BS

hort

ter

m o

utc

om

eLo

ng

-ter

m o

utc

om

e

Morr

is

200

8 (

26)

Extr

em

e lo

w

bir

th w

eig

ht

infa

nts

(<

100

0g

)

n =

1974

PT s

tart

d

irect

ly p

ost

nata

l ve

rsus

13

7 µm

ol/

L

(BW

50

1–75

0 g

) or

17

1 µm

ol/

L

(BW

751–

100

0g

)

Peak

TS

B hig

her

in c

onse

rvati

ve v

er-

sus

pro

phyl

act

ic g

roup

(16

8 v

ersu

s

120

µm

ol/

l, P<

0.0

01)

Dura

tion o

f PT

long

er in

pro

phyl

act

ic

vers

us

conse

rvati

ve g

roup

(88 v

ersu

s 35 h

rs P

<0

.00

1)

No s

ignific

ant

diffe

rence

in c

om

bin

a-

tion o

f d

eath

or

neuro

log

ica

l im

pair

-m

ent

Low

er r

isk

of n

euro

dev

elo

pm

enta

l im

pair

ment

in p

rop

hyla

ctic

gro

up

but

in t

his

gro

up

als

o a

NS

incr

ease

in m

or-

talit

y fo

r in

fants

with B

W 5

01–

750

g

Jang

aard

20

07

(25)

Low

bir

th

weig

ht

infa

nts

(<

150

0g

)

n =

95

PT s

tart

12

hours

post

nata

l ve

rsus

15

0 µ

mol/

L

No d

iffe

rence

in p

eak

TS

BN

o d

iffe

rence

in d

ura

tion o

f PT

Sub

gro

up

of

infa

nts

<10

00

g: P

eak

TS

B

hig

her

in t

he c

onse

rvati

ve v

ersu

s p

ro-

phyl

act

ic g

roup

(17

1 ve

rsus

139

µm

ol/

L,

P<0

.02)

No d

iffe

rence

s in

cer

eb

ral p

als

y and

/or

death

No d

iffe

rence

s in

menta

l dev

elo

pm

en-

tal i

ndex

Sub

gro

up

of

infa

nts

<10

00

g: N

S t

en-

dency

tow

ard

s p

oor

long

ter

m n

euro

-d

evelo

pm

enta

l outc

om

e in

conse

rva-

tive

gro

up

Curt

is-

Cohen

19

85 (

24)

Low

bir

th

weig

ht

infa

nts

(<

1250

g)

n =

22

PT s

tart

d

irect

ly p

ost

nata

l ve

rsus

85 µ

mol/

L

No d

iffe

rence

in p

eak

TS

BLo

wer

ag

e (

P<0

.00

1) a

nd lo

wer

TS

B

(P<0

.00

1) in

pro

phyl

act

ic g

roup

at

onse

t of

PT

Long

er d

ura

tion o

f PT

for

pro

phyl

act

ic

gro

up

ver

sus

conse

rvati

ve g

roup

(16

8

vers

us

121 hrs

, P<0

.05)

Not

desc

rib

ed

PT =

Photo

ther

ap

y, E

T =

Exc

hang

e T

ransf

usi

on, B

W =

Bir

th w

eig

ht, T

SB

= T

ota

l Ser

um

Bili

rub

in (

17.1

µmol/

L = 1 m

g/d

L b

iliru

bin

), N

S =

not

sig

nific

ant,

g =

gra

ms,

hrs

= h

ours

106

versus 120 µmol/L, conservative versus prophylactic phototherapy, respectively, p < 0.01). There was no difference between groups in the composite primary endpoint: a combination of death or neurodevelopmental impairment. The risk of neurodevel-opmental impairment was lower in the prophylactic group, but mortality was slightly higher, especially in preterm infants with a birth weight between 501 – 750 grams.(26) Unfortunately, we can not deduce ‘safe’ TSB-thresholds from these studies. Until now, there is no specific TSB concentration known at which phototherapy is more beneficial than harmful.(18)Next to variation in the TSB-thresholds, the present study showed differences in the criteria used to determine TSB-thresholds (gestational age versus birth weight) and in applied risk factors. Various risk factors for hyperbilirubinemia have been used the last 50 years. Risk factors reflect conditions associated with a high bilirubin production on the one hand, and conditions associated with bilirubin neurotoxicity on the other hand. Some examples of frequently used factors are: low birth weight, hypothermia, asphyxia, acidosis, hypoalbuminemia, sepsis, meningitis, intracranial haemorrhage and medication competing with bilirubin to bind albumin. Clinical evidence of most of these risk factors is limited; most factors are based on anecdotal clinical evidence or theoretical and experimental animal data.(1,3,10,27–34)  

Variability in management guidelines will for obvious reasons negatively influ-ence standardizing care for jaundiced newborn infants. Standardizing care based on the best available evidence, has been demonstrated to improve pediatric and adult patient outcomes.(35) Based on the best available evidence, i.e. the analysis of the guidelines used in the Dutch NICUs and on (inter)national guidelines, consensus was reached between the ten Dutch NICUs.(10,17)

The novel, ‘consensus based’ guideline is based on TSB-thresholds and includes five nomograms categorized in birth weight groups; < 1000 grams, 1000 – 1250 grams, 1250 – 1500 grams, 1500 – 2000 grams and > 2000 grams. TSB-thresholds for phototherapy and exchange transfusion are age-specific and depend on risk factors. TSB-thresholds at birth are not zero and rise to a plateau TSB-threshold 24–48 hours postnatal. Several reasons exist for this course of TSB-threshold concentrations. Data of TSB concentrations in umbilical cord blood show mean (±SD) TSB concentra-tions of 30 ± 9 µmol/L.(36) The slope of the curve which allows for treatment at lower TSB thresholds is chosen because: 1) bilirubin production is cumulative over time, 2) initially high TSB concentrations (exceeding TSB thresholds) inform the attending physician for the possibility of imminent severe hyperbilirubinemia (e.g. due to active hemolysis), and 3) bilirubin binding affinity for albumin is initially low which may be associated with higher free bilirubin concentrations.(37) Differences between standard or high risk infants are based on the commonly applied risk fac-

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 107

tors asphyxia (Apgar < 3 after 5 min), hypoxemia (PaO2< 5.3 kPa > 2 hour), acidosis (pH < 7.15 > 1 hour), active hemolysis (with positive Coombs), and clinical or neu-rological deterioration (such as sepsis with the need for vasopressants, meningitis or intracranial hemorrhage > grade 2 according to Papile).(38)

Other European guidelines for preterm infants are recently published in the UK and Norway.(15) In contrast to the Norwegian and Dutch guideline, the UK guideline is related to gestational age (and not birth weight) and includes specific nomograms for preterm infants of 23 to 37 weeks of gestational age.(16)

In The Netherlands, all preterm infants of less than 32 weeks of GA are admitted to a NICU. Therefore, this ‘consensus-based’ guideline was primarily developed for infants of 32 or less weeks of GA. Based on data of mean birth weights of newborn infants included in a Dutch national database (the Netherlands Perinatal Registry), 2 TSB-threshold nomograms of infants with a birth weight of 1500–2000 grams and more than 2000 grams seemed applicable for infants with gestational ages between 32 to 35 weeks.

In addition to the adapted American Academy of Pediatrics’ Subcommittee on Hyperbilirubinemia 2004 guideline for infants of 35 or more weeks of GA, a novel, ‘consensus-based’ guideline on unconjugated hyperbilirubinemia in the preterm infant is available. In 2008, the Dutch Society of Pediatrics approved and recommended this guideline for jaundiced infants of 35 or less weeks of GA.

Conclusion

Large variation in guidelines on hyperbilirubinemia existed in Dutch NICUs support-ing the world wide lack of evidence on treatment thresholds for jaundiced preterm infants. A novel, ‘consensus-based’ guideline for unconjugated hyperbilirubinemia in preterm infants of 35 or less weeks of GA has been developed. This novel guideline consists of TSB-based high and low risk treatment thresholds for five birth weight groups.

108

Acknowledgement

This project is funded by a grant from:

BARTrial studygroup: Academic Medical Center University of Amsterdam, the Netherlands:

Mrs. L. van Toledo-Eppinga, MD, PhD. University Medical Center Maastricht, the Netherlands:

A.L.M. Mulder, MD, PhD. Erasmus Medical Center Rotterdam, the Netherlands:

P. Govaert, MD, PhD. Isala Clinics Zwolle, the Netherlands:

R.A. van Lingen, MD, PhD. University Medical Center Leiden, the Netherlands:

E. Lopriore, MD, PhD. Maxima Medical Center Veldhoven, the Netherlands:

J. Buijs, MD. University Medical Center Groningen, the Netherlands:

Mrs. D.E. van Imhoff, MD; P.H. Dijk, MD, PhD; C.V. Hulzebos, MD, PhD. University Medical Center St. Radboud Nijmegen, the Netherlands:

K.D. Liem, MD, PhD. University Medical Center Utrecht, the Netherlands:

Mrs. M.J.N.L. Benders, MD, PhD. University Medical Center Amsterdam, the Netherlands:

W.P.F. Fetter, MD, PhD.

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 109

References

1. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003;88:F455-F458.

2. Oh W, Tyson JE, Fanaroff AA, et al. Association Between Peak Serum Bilirubin and Neurodevelopmental Outcomes in Extremely Low Birth Weight Infants. Pediatrics 2003;112:773–9.

3. American Academy of Pediatrics. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297–316.

4. Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 2005;25:54–9.

5. Bhutani VK, Maisels MJ, Stark AR, et al. Management of jaundice and prevention of severe neonatal hyperbilirubinemia in infants >or=35 weeks gestation. Neonatology 2008;94:63–7.

6. Dijk PH, de Vries TW, de Beer JJ. Guideline ‘Prevention, diagnosis and treatment of hyperbilirubinemia in the neonate with a gestational age of 35 or more weeks’. Ned Tijdschr Geneeskd 2009;153.

7. Watchko JF, Oski FA. Kernicterus in preterm newborns: past, present, and future. Pediatrics 1992;90:707–15.

8. Hansen TW. Therapeutic approaches to neonatal jaundice: an international survey. Clin Pediatr (Phila) 1996;35:309–16.

9. Rennie JM, Seghal A, De A, et al. Range of UK practice regarding thresholds for phototherapy and exchange transfusion in neonatal hyperbilirubinaemia. Arch Dis Child Fetal Neonatal Ed 2008.

10. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003;88:F459-F463.

11. Pearlman MA, Gartner LM, Lee K, et al. Absence of kernicterus in low-birth weight infants from 1971 through 1976: comparison with findings in 1966 and 1967. Pediatrics 1978;62:460–4.

12. Watchko JF, Claassen D. Kernicterus in premature infants: current prevalence and relationship to NICHD Phototherapy Study exchange criteria. Pediatrics 1994;93:996–9.

13. Maisels MJ. Clinical studies of the sequelae of hyperbilirubinemia. In: Levine RL, editor. Hyperbilirubinemia in the newborn, Report of the 85th Ross Conference on Pediatric Research.Columbus, OH: Ross Laboratories; 1983. p. 26–38.

14. Dani C, Poggi C, Barp J, et al. Current Italian practices regarding the management of hyperbilirubinaemia in preterm infants. Acta Paediatr 2011.

15. Bratlid D, Nakstad B, Hansen T. National guidelines for treatment of jaundice in the newborn. Acta Paediatr 2010.

16. National Institute for Health and Clinical Excellence. Neonatal Jaundice (Clinical guideline 98). 2010. www.nice.org.uk/CG98.

17. Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns. Pediatrics 1994;93:488–94.

110

18. Watchko JF, Jeffrey Maisels M. Enduring controversies in the management of hyperbilirubinemia in preterm neonates. Semin Fetal Neonatal Med 2010;15(3):136–40.

19. Shapiro SM. Bilirubin toxicity in the developing nervous system. Pediatr Neurol 2003;29:410–21.

20. Lucey JF. The unsolved problem of kernicterus in the susceptible low birth weight infant. Pediatrics 1972;49:646–7.

21. Lucey JF. Hyperbilirubinemia of prematurity. Pediatrics 1960;25:690–710.

22. Gartner LM, Snyder RN, Chabon RS, et al. Kernicterus: high incidence in premature infants with low serum bilirubin concentrations. Pediatrics 1970;45:906–17.

23. Newman TB, Maisels MJ. Evaluation and treatment of jaundice in the term newborn: a kinder, gentler approach. Pediatrics 1992;89:809–18.

24. Curtis-Cohen M, Stahl GE, Costarino AT, et al. Randomized trial of prophylactic phototherapy in the infant with very low birth weight. J Pediatr 1985;107:121–4.

25. Jangaard KA, Vincer MJ, Allen AC. A randomized trial of aggressive versus conservative phototherapy for hyperbilirubinemia in infants weighing less than 1500 g: Short- and long-term outcomes. Paediatr Child Health 2007;12:853–8.

26. Morris BH, Oh W, Tyson JE, et al. Aggressive vs. conservative phototherapy for infants with extremely low birth weight. N Engl J Med 2008;359:1885–96.

27. Lucey JF. Neonatal jaundice and phototherapy. Pediatr Clin North Am 1972;19:827–39.

28. Kim MH, Yoon JJ, Sher J, et al. Lack of predictive indices in kernicterus: a comparison of clinical and pathologic factors in infants with or without kernicterus. Pediatrics 1980;66:852–8.

29. Pearlman MA, Gartner LM, Lee K, et al. The association of kernicterus with bacterial infection in the newborn. Pediatrics 1980;65:26–9.

30. Turkel SB, Guttenberg ME, Moynes DR, et al. Lack of identifiable risk factors for kernicterus. Pediatrics 1980;66:502–6.

31. Van de Bor M, Ens-Dokkum M, Schreuder AM, et al. Hyperbilirubinemia in low birth weight infants and outcome at 5 years of age. Pediatrics 1992;89:359–64.

32. Levine RL, Fredericks WR, Rapoport SI. Entry of bilirubin into the brain due to opening of the blood-brain barrier. Pediatrics 1982;69:255–9.

33. Ebbesen F, Brodersen R. Risk of bilirubin acid precipitation in preterm infants with respiratory distress syndrome: considerations of blood/brain bilirubin transfer equilibrium. Early Hum Dev 1982;6:341–55.

34. Satar M, Atici A, Oktay R. The influence of clinical status on total bilirubin binding capacity in newborn infants. J Trop Pediatr 1996;42:43–5.

35. Muething SE. Improving patient outcomes by standardizing care. J Pediatr 2005;147:568–70.

36. Knupfer M, Pulzer F, Gebauer C, et al. Predictive value of umbilical cord blood bilirubin for postnatal hyperbilirubinaemia. Acta Paediatr 2005;94:581–7.

Uniform treatment thresholds for hyperbilirubinemia in preterm infants 111

37. Ebbesen F, Nyboe J. Postnatal changes in the ability of plasma albumin to bind bilirubin. Acta Paediatr Scand 1983;72:665–70.

38. Papile LA, Burstein J, Burstein R, et al. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr 1978;92:529–34.

56

9SD

7

234

8

113

Chapter 7

High variability and low irradiance of

phototherapy devices in Dutch NICUs

Deirdre E. van Imhoff, Christian V. Hulzebos, Maaike van der Heide, Vera W. van den Belt,

Henk J. Vreman, Peter H. Dijk,

and the BARTrial Study Group

Arch Dis Child Fetal Neonatal Ed. 2012 May 18. (Epub ahead of print)

114

Abstract

Objective: To evaluate phototherapy practices by measuring the irradiance levels of phototherapy (PT) devices.

Design: Prospective study

Setting: Tertiary neonatal intensive care units

Patients: None

Interventions: Irradiance levels of PT devices used in the ten Dutch NICUs were measured according to the local PT practice patterns. The irradiance levels of all overhead and fibre optic PT devices were measured with a radiometer using an infant silhouette model.

Results: Eight different PT devices were used in the ten NICUs; five were overhead devices and three fibre optic pads. The median (range) irradiance level for overhead PT devices was 9.7 (4.3 – 32.6) µW/cm2/nm and for fibre optic pads 6.8 (0.8 – 15.6) µW/cm2/nm. Approximately 50% of PT devices failed to meet the minimal recom-mended irradiance level of 10µW/cm2/nm. Maximal irradiance levels for overhead PT spot lights were inversely related to the distance between device and infant model (R2=0.33). The distances ranged from 37 to 65 cm.

Conclusions: PT devices in the Dutch NICUs show considerable variability with of-ten too low irradiance levels. These results indicate that suboptimal PT is frequently applied and may even be ineffective towards reducing TSB levels. These results underline the need for greater awareness among all health care workers towards the requirements for effective PT including measurements of irradiance and distance.

High variability and low irradiance of phototherapy devices in Dutch NICUs 115

Introduction

Unconjugated hyperbilirubinaemia is very common in newborn infants and may result in bilirubin neurotoxicity. PT is an effective and safe treatment for reducing total serum bilirubin (TSB).(1)

Specific TSB thresholds to start PT treatment have been defined in several in-ternational guidelines.(2–4) These guidelines contain also recommendations for the effective irradiance level of PT, which is strongly related to the decrease in TSB.(2,5) The irradiance level within the effective bilirubin photo degradation spectrum (400–520 nm with a peak at 460 nm) to a sufficient treatment area can be measured with a radiometer. The American Academy of Pediatrics (AAP) recommends that effective PT should occur with a minimal irradiance level of 8 to 10 µW/cm²/nm.(2,6) Adherence to this recommendation has not been investigated thoroughly. Only a few studies from developing countries measured low irradiance suggesting that PT is not always applied effectively.(7–9) Limited resources with inferior light sources, maintenance constraints, inconsistent power supply and lack of awareness of health care workers are probably the most important barriers to effective PT.(10) We hy-pothesize that in the technically advanced NICUs in industrialised countries these barriers do not exist and that PT is applied more effectively. Therefore we measured the irradiance levels of PT devices in the ten Dutch level III Neonatal Intensive Care Units (NICUs). 

Methods

All ten Dutch level III NICUs were invited to be visited for measurements of irradi-ance levels under practice conditions. At least one sample of each type of PT device in combination with each type of incubator in use in each NICU was included.

The PT devices that were studied included overhead PT devices with diffuse light or spot patterns illumination using fluorescent tube or halogen lamps, and underneath or contact devices using halogen lamps with fibre optic cables and pads (Table 1).

116

Tabl

e 1. P

hoto

ther

apy d

evic

es, ir

radi

ance

leve

ls an

d di

stan

ces.

PT d

evic

eN

ICU

s n

Irra

dia

nce

pro

vid

ed

by

MD

A o

r m

anufa

cture

r ra

ng

es

(µ

W/c

m2 /

nm

)#

Irra

dia

nce

m

easu

red

m

ed

ian (

rang

e) n

(µ

W/c

m2 /n

m)

Dis

tance

re

com

mend

ed

by

manufa

cture

r (c

m)

Dis

tance

m

easu

red

med

ian (

rang

e)

(cm

)

Ove

rhead

dev

ices

:

Ohm

ed

a S

pot

Lam

p1

852–

788.6

(4

.6–1

4.0

) 18

38–6

04

6 (

37–6

5)

Gir

aff

e S

pot

Lite

23

29–5

819

.5 (

18.1–

19.7

) 4

38–6

04

6 (

38

–53)

Med

ela

Lam

p3

114

–31

9.6

(8.4

–10

.8)

225

–45

40

(38

–42)

Drä

ger

Unit

40

00

42

14–2

721

.4 (

8.3

–32.

6)

530

–40

50

(38

–50

)

Fibre

Op

tic p

ad d

evic

es:

Ohm

ed

a li

ght5

315

–35

5.9

(1.8

–6.4

) 3

––

Ohm

ed

a p

lus5

632

6.5

(0

.8–1

4.0

) 6

––

Ohm

ed

a h

igh

54

15–3

514

.9 (

9.6

–15.6

) 4

––

1. Date

x-O

hm

ed

a P

T S

pot

Lam

p (

H):

Ohm

ed

a M

ed

ica

l, C

olu

mb

ia, M

D, U

SA

, 2.

Gir

aff

e S

pot

PT L

ite (

H):

Gener

al E

lect

ric

Hea

lthca

re, L

aure

l, M

D, U

SA

, 3. M

ed

ela

PT

lam

p (

F):

Med

ela

AG

Med

ica

l Tech

nolo

gy,

Baar,

Sw

itze

rland

, 4

. Drä

ger

PT

Unit

40

00

(F):

Drä

ger

Med

ica

l, Lü

beck

, Ger

many,

5. D

räg

er-H

eraeus

PT U

nit

80

0 (

GD

): D

räg

er M

ed

ica

l, Lü

beck

, Ger

many,

6

. Ohm

ed

a (

H a

nd F

O):

Ohm

ed

a M

ed

ica

l, C

olu

mb

ia, M

D, U

SA

. Ty

pe P

T la

mp

: H =

ha

log

en, F

= f

luore

scent, G

D =

ga

llium

dis

charg

e b

ulb

and

FO

= f

ibre

op

tic

# Ir

r ad

iance

was

calc

ulate

d f

rom

data

pro

vid

ed

by

the M

ed

ica

l Dev

ice A

gency

(M

DA

) b

y d

ivid

ing

irra

dia

nce

in m

W/c

m2 b

y th

e s

pect

ral r

ang

e (

whic

h

was

40

0–5

50

nm

= 150

), t

he lo

west

and

hig

hest

va

lues

are

pre

sente

d.(

11–1

5)

High variability and low irradiance of phototherapy devices in Dutch NICUs 117

A silhouette model representing a preterm infant was used to measure the irradiance levels of overhead PT devices. Five measurement points in craniocaudal direction were marked on the model representing the head, trunk, abdomen, knees and feet (3, 12, 18, 23 and 33 cm, respectively, figure 1). Local NICU nurses placed the silhouette model in the incubator and installed the PT device above or on top of the incubator as they would do in daily clinical practice. The type of PT device, type of incubator, the distance between PT device and silhouette model were noted. The irradiance levels of underneath contact fibre optic PT pads were measured on top of the rou-tinely used disposable cover at three predefined measure points (12, 18 and 23 cm respectively, figure 1). 

Figure 1. Silhouette models with measurement points used for irradiance level measurements of overhead PT devices and underneath fibre optic PT devices.

A. Irradiance levels of panel and spot overhead PT devices were measured at the five indicated points in craniocaudal direction on a 35 cm long silhouette model representing a preterm infant.

B. For underneath fibre optic PT devices, irradiance was measured at the 3 indicated points on a disposable cover.

118

A Dale40 Phototherapy Radiometer (Fluke Biomedical, Everett, WA, USA) was used to measure the energy distribution. This radiometer is designed to measure light radiation of fluorescent lamps in the blue part of the spectrum with a band range of 429 to 473 nm and peak sensitivity at 453 nm. The measured energy divided by the bandwidth (width of sensitivity spectrum at 50% of maximum) of 44 nm as is reported by the manufacturer resulted in the integrated irradiance (µW/cm2/nm) delivered by the device at the measured distance from the light exit point.

Five replicate irradiance measurements were performed at each of the five or three measurement points (as appropriate) for each PT device studied. The mean values of the five replicates were used for further calculations. For each PT device type tested, the means of all measurement points were averaged to one mean ir-radiance level, in agreement with the definition of the ‘effective surface area’ by the International Electrotechnical Commission and the recommendation of the AAP.(2) For each NICU, the mean irradiance levels per device tested were used to calculate medians, quartiles and outliers to be represented in box plots. To evaluate the effect of distance on irradiance of PT spot devices, the irradiances of the measurement points the closest to the devices were used (which are the measurement points with the highest irradiance levels). 

Microsoft Office Excel (Microsoft Corporation, Redmond, Washington) and SPSS for Windows (version 16.0, Chicago, IL) were used for data entry and analysis. 

Statistical tests: Alternate T-test was used for comparing the irradiance on the five measurement points on the model for PT tube lights versus PT spot lights. A p value < 0.05 was considered statistically significant.

Results

All ten level III Dutch NICUs participated. Eight types of PT devices were found to be in use; five types of overhead PT devices and three types of underneath fibre optic PT pads (table 1). Five types of incubators were used (Giraffe, Dräger Caleo, Dräger 8000, Vita and Hill Rom Airshield). We measured a total of 42 PT device – incubator combinations (29 overhead and 13 fibre optic devices).

Figure 2 shows the variation in irradiance in each NICU of all PT devices together. The irradiance levels ranged from 0.8 to 32.6 µW/cm2/nm.

High variability and low irradiance of phototherapy devices in Dutch NICUs 119

0

10

20

30

40

1

NICU

Irra

dia

nce

(µW

/cm

2 /nm

)

2 3 4 5 6 7 8 9 10 All

Figure 2. Irradiance levels of all phototherapy devices in the ten NICUs. Irradiance levels for all PT devices (n=42) in each of the ten NICUs and for all NICUs together. The median is marked by the horizontal line in the box. The boxes are limited by the 25th and 75th percentile. The whiskers (┴) represent the lowest and highest irradiance levels measured within 1.5 interquartile distance below or above the box. Outliers (○) are depicted separately and represent irradiance levels between 1.5 and 3 interquartile distance below or above the box. The minimal recommended irradiance level of PT (8 – 10 µW/cm2/nm) is represented by the grey horizontal line.

Figure 3 shows the ranges in irradiance levels for all the overhead PT devices and fibre optic pads. The median (range) irradiance level for overhead PT devices was 9.7 (4.3 – 32.6) µW/cm2/nm and for fibre optic pads 6.8 (0.8 – 15.6) µW/cm2/nm. Irradiance levels were lower than 10 µW/cm2/nm in 52% of the overhead and 69% of the fibre optic underneath PT devices. Table 1 shows the median irradiance level for each type of PT device and the calculated irradiance as provided by the manufacturer and/or Medical Device Agency, together with the recommended and measured distances.(11–15)

120

0

10

20

30

40

Overhead PT Underneath PT

Irra

dia

nce

(µW

/cm

2 /nm

)

Figure 3. Irradiance levels of overhead PT and underneath fibre optic PT devices. Irradiance levels of overhead PT (n=29) compared to underneath fibreoptic PT (n=13) devices. The median is marked by the horizontal line in the box. The boxes are limited by the 25th and 75th percentile. The whiskers (┴) represent the lowest and highest irradiance levels measured within 1.5 interquartile distance below or above the box. Outliers (○) are depicted separately and represent irradiance levels between 1.5 and 3 interquartile distances below or above the box. The minimal recommended irradiance level of PT (8 – 10 µW/cm2/nm) is represented by the grey horizontal line.

The distance between PT-device and silhouette model ranged from 37 to 65 cm. Figure 4 shows the significant relationship (R2=0.33) between the increasing distance and lower irradiance levels of PT spots included in this study. We found no significant differences between the distances at which the PT devices were positioned per type of incubator or type of PT device (data not shown).

High variability and low irradiance of phototherapy devices in Dutch NICUs 121

0

10

20

30

40

30

y=90.826 e–0.031x

R2=0.3256

Distance between PT device and silhouette model (cm)

Irra

dia

nce

(µW

/cm

2 /nm

)

35 40 45 50 55 60 65

Figure 4. The relationship between irradiance level and distance between PT device and silhouette model of Spot PT devices. Irradiance level corresponding to the shortest distance between the PT device and the measurement point on the silhouette model (and consequently the highest irradiance) for the Ohmeda Spot lights and Giraffe Spot PT Lite lights (n=22). A statistical significant natural logarithmic correlation with a R2 of 0.33 was found between irradiance (y) and distance (x): y=90.826e -0.031x. Distances ranged from 37 to 65 cm and irradiance levels from 7.4 to 41.2 µW/cm2/nm.

Figure 5 shows the mean irradiance level per measurement point on the silhouette model for tube lights compared to spot lights. Irradiance levels were higher for the tube – versus spot PT devices on measurement points 1, 4 and 5 (p=0.058, p=0.03 and p=0.005, respectively), but not on measurement points 2 and 3, indicating a smaller footprint for spot-devices.

122

0

10

20

30

40

1

Measurement point on silhouette model

Irra

dia

nce

(µW

/cm

2 /nm

)

2 3 4 5

Tube lights

Spot lights

p=0.005

p=0.03

p=0.058

Figure 5. Irradiance level per measurement point on the silhouette model for tube and spot overhead PT devices. Irradiance levels for each measurement point on the silhouette model for tube PT devices n=6 (Dräger PT Unit 4000 and Medela PT lamps) compared to spot PT devices n=22 (Ohmeda Spot PT lamp and Giraffe Spot PT Lite). Data represent mean values and standard deviations. P values are the result of alternate T-tests comparisons between tube and spot PT devices.

Discussion

We found a wide variation in irradiance levels of the PT devices in daily clinical use in Dutch NICUs. Approximately 50% of the devices delivered irradiance levels lower than the minimal recommended level of 10 µW/cm2/nm.(2,6) These results indicate that suboptimal PT is often applied and may even be ineffective towards reducing TSB levels. We established that this low irradiance was due to excessive distances between PT device and infant and disappointing performance of PT devices.

High variability and low irradiance of phototherapy devices in Dutch NICUs 123

We had expected that in NICUs in an industrialised country such as the Nether-lands, PT would often meet the irradiance level that is recommended for effective PT. However, we found that it does not. In fact, our results are not that different from those in developing countries.(7–9) Pejaver et al. studied PT devices among 24 Indian neonatal care centres. Only 31% of 58 PT devices provided an acceptable irradiance level.(7) Similarly, Ferreira et al. found that only 33.3% of the 36 PT devices in Maceió, Brazil, provided an irradiance level of 10 µW/ cm2/nm or more.(8) Owa et al. studied 63 PT devices at 12 nurseries in Nigeria. Only 6% provided an irradiance level of 10 µW/cm2/nm or more, and 75% of less than 5 µW/cm2/nm.(9)

Bhutani et al. described four categories of barriers to effective PT in developing countries that may explain the high percentage of PT devices that produce subop-timal levels of irradiance. Firstly, PT devices may contain inferior light sources with suboptimal spectral range. Secondly, maintenance constraints may result in devices with burned out or missing lights. Thirdly, environmental barriers such as lack of consistent electrical power supply may negatively influence the efficacy of PT. The fourth barrier consists of a lack of awareness of requirements for effective PT among the health care workers that administer PT, and the lack of radiometers to assess the efficacy of PT.(10) While resource-constrainment is probably the most important contributor in the developing countries, lack of awareness of the factors that influ-ence the effectiveness of PT is probably the most important barrier in industrialised countries.

The first factor is the PT device itself and its specifications such as light source, intensity, spectral range and lifetime.(10,16) The light sources vary from fluorescent to halogen lights, gas discharge tubes and gallium nitride light emitting diodes (LED). The number of light bulbs and colour of the light also differ between PT devices. Vari-ous PT devices are used in the Dutch NICUs (table 1). Since these irradiance levels were measured not in a standardized laboratory setting, but in the clinical setting, our results may be different from those provided by manufacturers and agencies that evaluate medical devices like the MDA.(11–14) We think that a clinical evaluation, like ours, may be more veracious than standardized laboratory evaluations.

Maintenance and especially lifetime of the light sources are important for the performance of PT devices, but these were not systematically evaluated in this study. However, NICU nurses were often not aware of the lifetime of the lights, which may have been more prone to suboptimal functioning than expected.

The second factor influencing the effectiveness of PT is the distance between the PT device and the infant. Positioning the PT device as close to the infant as possible increases irradiance levels and consequently the efficacy of PT (figure 4).(6) However, if a specific light source produces too much heat, as could happen with halogen-lamp

124

based devices, close positioning of the lamp to the infant may increase the risk of heat burn.(2,17) Therefore, it is very important to follow the safety instructions of the manufacturer.

In the present study, several clinical variables may have influenced the distance between the infant and the PT devices. Five types of incubators were used, all with different dimensions. Although the height of an incubator limits the minimal distance between PT device and the infant, the type of incubator was not the most important variable that determined the operational distance. We found a broad range of applied distances for each type of incubator, and no clear relationship between the type of incubator and distance or irradiance levels. Apparently, the positioning of the PT devices above the incubator by the NICU nurses was different and contributed in particular to the differences in distances.

The third factor that influences the effectiveness of PT is the amount of surface area of naked skin of the infant that is exposed to a PT light. Firstly, baby hats, diapers and clothes reduce the surface area of naked skin exposed to the light. Furthermore, the positioning of the PT light above the incubator or the degree of dispersion be-tween the light source and the infant affects irradiance. Figure 5 illustrates that the footprints of spot lights have considerably higher irradiance levels in the centre of the light and much less intense irradiance levels at the periphery compared to tube devices.(18) This figure also illustrates that improper positioning of the PT device can seriously reduce irradiance levels.(17)

Our primary goal was not to compare the performance of different types of PT devices, but to compare clinical practice conditions of PT. As such, we chose to mea-sure irradiance with one type of radiometer. PT devices differ in delivered spectrum of light while radiometers differ in the measured spectrum of light. Manufacturers of PT devices recommend the use of radiometers that fit the spectrum of the light deliv-ered by their device. The ideal handheld radiometer that would equally measure the complete and only treatment spectrum of PT light does not yet exist. The radiometer that we used measured irradiance levels with a bandwidth between 429 to 473 nm, which is within the effective treatment bandwidth of 430–490 nm, recommended by the AAP.(2) Irradiance levels are expressed as µW/cm2/nm, which means that the absolute results of the measurements depend on the bandwidth and sensitivity level of the radiometer.(16,19) The irradiance levels of this study may differ from those of other studies or specifications given by the manufacturers of PT devices (table 1). This underlines the need for a universal radiometer.(19)

High variability and low irradiance of phototherapy devices in Dutch NICUs 125

Conclusion

This study shows that the irradiance levels of PT devices in level III NICUs in an industrialised country, such as the Netherlands, are not that different from subop-timal irradiance levels of PT devices observed in developing countries. Resource-constrained related barriers are probably not the most important in industrialised countries. Awareness and understanding of the factors that influence the efficacy of PT by all health care professionals that buy, apply and maintain PT devices, seems of key importance (table 2). In particular, the distance between PT device and in-fants turned out to be an important factor toward optimising irradiance. Frequently measuring the irradiance levels of PT devices is essential, and using only devices that deliver effective PT under clinical conditions. Finally, our results underline the need for a universal radiometer that measures the clinical relevant irradiance level of all PT devices.

Table 2. Recommendations for effective use of phototherapy

Recommendations

Educate all health care professionals who buy, apply and maintain PT devices about the factors that affect the efficacy of PT.

Measure the irradiance level of your PT devices regularly. The minimal recommended irradiance level is 8 to 10 µW/cm2/nm. Replace lamps that do not deliver.

Match PT devices to incubators in use and place the PT device as close to the infant as is safe and possible. Follow the safety instructions of the manufacturer of the PT device to avoid heat and burns (especially with halogen lights)!

Illuminate as much naked skin of the infant as possible, making sure optimal body tem-perature is maintained.

Use eye protection.

126

Funding

This study preceded the Netherlands Neonatology Research Network RCT ‘Reducing bilirubin induced neurological dysfunction in premature newborns: additional use of the bilirubin/albumin ratio in the treatment of hyperbilirubinemia’ (BARTrial: ISRCTN 7446543). The trial was funded by the ZonMW Cost-Effectiveness programme (nr: 94507407)  

Acknowledgements

The BARTtrial Study Group members are: Academic Medical Center University of Amsterdam, the Netherlands:

L. van Toledo-Eppinga, MD, PhD. University Medical Center Maastricht, the Netherlands:

A.L.M. Mulder, MD, PhD. Erasmus Medical Center Rotterdam, the Netherlands:

P. Govaert, MD, PhD. Isala Clinics Zwolle, the Netherlands:

R.A. van Lingen, MD, PhD. University Medical Center Leiden, the Netherlands:

E. Lopriore, MD, PhD. Maxima Medical Center Veldhoven, the Netherlands:

J. Buijs, MD. University Medical Center Groningen, the Netherlands:

D.E. van Imhoff, MD; P.H. Dijk, MD, PhD; C.V. Hulzebos, MD, PhD. University Medical Center St. Radboud Nijmegen, the Netherlands:

K.D. Liem, MD, PhD. University Medical Center Utrecht, the Netherlands:

M.J.N.L. Benders, MD, PhD. University Medical Center Amsterdam, the Netherlands:

W.P.F. Fetter, MD, PhD.

High variability and low irradiance of phototherapy devices in Dutch NICUs 127

References

1. Maisels MJ. Phototherapy–Traditional and non-traditional. J Perinatol 2001;21:S93–7.

2. American Academy of Pediatrics. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297–316.

3. Rennie J, Burman-Roy S, Murphy MS. Neonatal jaundice: summary of NICE guidance. BMJ 2010;340:c2409.

4. Van Imhoff DE, Dijk PH, Hulzebos CV. Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a national guideline. Early Hum Dev 2011;87:521–5.

5. Tan KL. The pattern of bilirubin response to phototherapy for neonatal hyperbilirubinaemia. Pediatr Res 1982;16:670–4.

6. Maisels MJ. Why use homeopathic doses of phototherapy? Pediatrics 1996;98:283–7.

7. Pejaver RK, Vishwanath J. An audit of phototherapy units. Indian J Pediatr 2000;67:883–4.

8. Ferreira AL, Nascimento RM, Verissimo RC. Irradiance of phototherapy equipment in maternity wards in Maceio. Rev Lat Am Enfermagem 2009;17:695–700.

9. Owa JA, Adebami OJ, Fadero FF et al. Irradiance Readings of Phototherapy Equipment: Nigeria. Indian J Pediatr 2011;78:996–8.

10. Bhutani VK, Cline BK, Donaldson KM et al. The need to implement effective phototherapy in resource-contrained setting. Sem Perinatol 2011;35:192–7.

11. Medical Devices Agency: Neonatal Phototherapy Datex Ohmeda Spot PT Lamp. MDA Evaluation 00092, London, GB, May 2001. http://www.wales.nhs.uk/sites3/Documents/443/00092%20Datex-Ohmeda%20Spot%20Phototherapy%20Lamp.pdf.

12. Medical Devices Agency: Neonatal Phototherapy. Draeger PT 4000 Unit. MDA Evaluation 01162, London, GB, December 2001. http://www.wales.nhs.uk/sites3/Documents/443/01162%20Draeger%20Phototherapy%204000%20Unit.pdf.

13. Medical Devices Agency: Neonatal Phototherapy. Medela Phototherapy Lamp. MDA Evaluation 01161, London, GB, December 2001. http://www.wales.nhs.uk/sites3/Documents/443/01161%20Medela%20Phototherapy%20Lamp.pdf.

14. Medical Devices Agency: Neonatal Phototherapy. A review including evaluations of Ohmeda Biliblanket Plus and Medela BiliBed. MDA Evaluation 391, London, GB, April 2000. http://www.wales.nhs.uk/sites3/Documents/443/391%20Neonatal%20Phototherapy%20Review.pdf.

15. Giraffe Spot PT Lite Phototherapy System Brochure 2007. GE Healthcare.    http://www.gehealthcare.com/euen/maternal-infant-    care/docs/Giraffe_SPOT_PT_Lite_bro_e.pdf.    

16. Maisels MJ, McDonagh AF. Phototherapy for neonatal jaundice. N Engl J Med 2008;358:920–8.

17. Stokowski LA. Fundamentals of phototherapy for neonatal jaundice. Adv Neonatal Care 2006;6:303–12.

128

18. Vreman HJ, Wong RJ, Murdock JR et al. Standardized bench method for evaluating the efficacy of phototherapy devices. Acta Paediatr 2008;97:308–16.

19. Vreman HJ. Phototherapy: the challenge to accurately measure irradiance. Indian J Pediatr 2010;47:127–8.

23456789SD

129

Chapter 8

Measurements of neonatal bilirubin and

albumin concentrations: A need for

improvement and quality control

Deirdre E. van Imhoff, Peter H. Dijk, Cas W. Weykamp, Christa M. Cobbaert, Christian V. Hulzebos

On behalf of the BARTrial studygroup

European Journal of Pediatrics. 2011;170:977–82.

130

Abstract

Background: Accurate and precise bilirubin and albumin measurements are es-sential for proper management of jaundiced neonates. Data hereon are lacking for Dutch laboratories.

Objective: We aimed to determine variability of measurements of bilirubin and albumin concentrations typical for (preterm) neonates.

Methods: Aqueous, human serum albumin based samples with different concentra-tions of bilirubin (100, 200, 300, 400, and 500 µmol/L) and albumin (0, 10, 15, 20, 25 and 30 g/L) were sent to laboratories of all Dutch neonatal intensive care units (n=10). Bilirubin and albumin recoveries of the specimens were measured using locally available routine analytical methods. The mean, standard deviation (SD), and coefficients of variations (CV) were calculated per sample.

Results: Bilirubin concentrations were underestimated in the absence of albumin (maximal CV: 26.0%). When the albumin concentration was 10 or 20 g/L, the bilirubin concentrations of the samples were overestimated (maximal CV 14.1 and 9.2%, resp.). Variability increased with higher weighed-in bilirubin concentrations. Measured albumin levels were ~10% lower than albumin levels of manufactured samples. Bilirubin concentration did not influence albumin measurements. The maximal CV was 6.8%.

Conclusion: Interlaboratory variability of bilirubin and albumin measurements is high. Recalibration and introduction of a specific quality assessment scheme for neonatal samples is recommended to ensure exchangeability of bilirubin and albumin measurements among laboratories and to control the observed large variability.

Measurements of neonatal bilirubin and albumin concentrations 131

Introduction

The risk of severe hyperbilirubinemia and bilirubin neurotoxicity in newborn infants underlines the need for accurate and precise bilirubin and albumin measurements as a part of management guidelines for hyperbilirubinemia. To prevent severe uncon-jugated hyperbilirubinemia and bilirubin neurotoxicity, (inter)national guidelines are used to standandize management of jaundiced neonates.(1,3,9,14) Treatment of unconjugated hyperbilirubinemia is based on the measurement of total serum bilirubin (TSB) concentration. Measurement of albumin concentration is recom-mended because a low albumin concentration is considered a risk factor for bilirubin neurotoxicity, resulting in lower TSB treatment thresholds or albumin infusion.(1)

Variability of bilirubin and albumin measurements will obviously affect treatment decisions. In general, variability of routine chemical parameters is controlled by Qual-ity Assessment Schemes using External Quality Assessment (EQA) samples. In The Netherlands, human matrix based, commutable EQA-samples provided by the Dutch Foundation for Quality Assessment in Clinical Laboratories (in Dutch: SKML) are biweekly analyzed. However, bilirubin and albumin concentrations of these EQA-samples typically represent adult ranges with a maximal bilirubin concentration of 80 µmol/L and a minimal albumin concentration of 30 g/L. Higher concentrations of bilirubin and lower concentrations of albumin are frequently encountered in (preterm) neonates. Data on variability of bilirubin and albumin. measurements in the neonatal range are scarce and unavailable for Dutch laboratories.(7) In this study, we aimed to assess variability in measurements of bilirubin and albumin concentra-tions typical for neonates.

Methods

Participating laboratories

To analyze the variability in bilirubin and albumin measurements, the Dutch SKML distributed samples to laboratories of all (n=10) Dutch Neonatal Intensive Care Units (NICUs). The NICUs were included because NICUs – and not general hospitals – are involved in the care of preterm infants of 32 weeks or less of gestational age having an increased risk of developing hyperbilirubinemia and bilirubin neurotoxicity.

Specimens

The Dutch SKML manufactured artificial, aqueous, albumin based EQA-samples to evaluate bilirubin and albumin assay performance in the neonatal concentration range. In a TRIS-HCl buffer (pH 7.4) with 9 g/L NaCl, human serum albumin (HSA)

132

(Sigma A 1653) was dissolved to achieve albumin concentrations ranging from 0 – 30 g/L (0, 10, 15, 20, 25 and 30 g/L). Unconjugated bilirubin (Sigma B4146) was dis-solved (protected from light) and added to the albumin solutions immediately after complete solution. Bilirubin concentrations ranged from 100 to 500 μmol/L (100, 200, 300, 400, and 500 µmol/L). Table 1 shows the combinations of bilirubin and albumin concentrations in the samples. After manufacture the samples were frozen below -70˚C, shipped on dry-ice to the participating laboratories and stored frozen until analysis. All EQA-samples were analyzed in April 2008.

Table 1. Combinations of albumin and bilirubin concentrations in manufactured EQA-samples

Sample 1 2 3 4 5 6 7 8 9 10

Albumin (g/L)

0 0 0 0 0 10 10 10 10 10

Bilirubin (mmol/L)

0 100 200 300 400 0 100 200 300 400

Sample 11 12 13 14 15 16 17 18 19 20

Albumin (g/L)

10 15 20 20 20 20 20 20 25 30

Bilirubin (mmol/L)

500 0 0 100 200 300 400 500 0 0

17.1 µmol/L = 1 mg/dL bilirubin

Analytical devices

In the clinical chemistry laboratories of the corresponding NICUs (n = 10), bilirubin and albumin recoveries of the processed EQA-samples were evaluated using locally available routine analytical methods. Measurements on Point of Care analyzers were not included. All routine methods were standardized according to the instructions of the manufacturers. All laboratories participate in the regular EQA-program of the SKML with maximal bilirubin concentrations of 80 µmol/L and minimal albumin concentrations of 30 g/L, enabling them to monitor and control albumin and bili-rubin performance in the normal and slightly abnormal concentration ranges. The majority of the samples was measured in triplicate; a minority (1 center) in duplicate.

Statistical analysis

The weighed-in concentrations of bilirubin and albumin in the processed EQA-samples were compared with the measured concentrations as recovered by the clinical

Measurements of neonatal bilirubin and albumin concentrations 133

chemistry laboratory using routine methods. The mean, standard deviation (SD), range and coefficients of variations (CV) were calculated per measured sample using Microsoft Office Excel (Microsoft corporation, Redmond, Washington) and SPSS for Windows (version 16.0, Chicago, IL, USA). 

Results

Analyzers and methods

Bilirubin and albumin measurements in the NICU laboratories were performed with locally available analytical methods, applied on routine clinical chemistry analyzers. Table 2 shows the analytical devices and corresponding methods used in the NICU laboratories.

Table 2. Analytical devices and methods used in Dutch NICU laboratories

Devices used for bilirubin measurement Method Number of NICUs

Roche Modular DPD* 6

Reichert Unistat bilirubinometer** spectrophotometric 1

Bilimeter 3*** spectrophotometric 1

Beckman Coulter Jendrassik Grof 1

Abbot Aeroset Jendrassik Grof 1

Devices used for albumin measurement Method Number of NICUs

Roche Modular 7

P800 module Bromcresol green 6

Automatic Bromcresol purple 1

Beckman Coulter 2

Synchron Lx20 Bromcresol green 1

DxC 800 Bromcresol purple 1

Abbott Aeroset Bromcresol purple 1

* DPD = 2,5-DiCl-Phenyl Diazonium, ** Unistad Salm and Kipp BV, *** Pfaff Technik & Medizin

Bilirubin

Figures 1A, B, and C show the measured versus the weighted bilirubin concentra-tions in samples with albumin concentrations of 0, 10 and 20 g/L, respectively. Underestimation of bilirubin concentrations occurred in the absence of albumin (Figure 1A). When the albumin concentration was 10 or 20 g/L (Figure 1B and C),

134

the bilirubin concentrations of the samples were overestimated. Independent of the albumin concentration in the sample, a large interlaboratory variability was found. The interlaboratory variability increased with increasing bilirubin concentrations (Figure 1B and C). For example, for a bilirubin concentration of 100 µmol/L and an albumin concentration of 10 g/L, the range in measured bilirubin concentration was 97 to 128.5 µmol/L (31.5 µmol/L), ~30% of the concentration in the sample (Figure 1 B). For bilirubin concentrations with an albumin concentration of 20 g/L, this increase in interlaboratory variability was less (Figure 1C). Maximal interlaboratory coefficients of variation of measured bilirubin concentrations were 26.0, 14.1 and 9.2% for samples with albumin concentrations of 0, 10 and 20 g/L, respectively.

00

100

200

30

0

40

0

100

Bilir

ub

in (

µmol/

L) in

manufa

cture

d c

ontr

ol

Measured bilirubin (µmol/L)

200

30

04

00

AA

lbum

in 0

g/L

Measurements of neonatal bilirubin and albumin concentrations 135

00

100

200

30

0

40

0

50

0

60

0

100

200

40

0

Bilir

ub

in (

µmol/

L) in

manufa

cture

d c

ontr

ol

Measured bilirubin (µmol/L)

30

050

06

00

BA

lbum

in 10

g/L

CV

(%)

100

200

30

0

40

0

50

0

31.5

22.0

19.5

39

.8

48.4

9.4

6.8

6.7

12.1

14.1

Measu

red

bili

rub

in( µ

mol/

L)

Min

-max

diffe

rence

(% m

anu-

fact

ure

db

iliru

bin

)

00

100

200

30

0

40

0

50

0

60

0

100

200

40

0

Bilir

ub

in (

µmol/

L) in

manufa

cture

d c

ontr

ol

Measured bilirubin (µmol/L)

30

050

06

00

CA

lbum

in 2

0 g

/L

100

200

30

0

40

0

50

0

29.0

25.8

23.8

24.1

19.3

9.2 8.1

7,3

7.9

6.6

Measu

red

bili

rub

in( µ

mol/

L)

CV

(%)

Min

-max

diffe

rence

(% m

anu-

fact

ure

db

iliru

bin

)

Figu

re 1A

–C. V

aria

bilit

y of b

iliru

bin

mea

sure

men

ts.

Fi

gure

s 1A

, 1B,

and

1C sh

ow m

easu

red

vers

us w

eigh

ed-in

bili

rubi

n co

ncen

tratio

ns (1

7.1 µ

mol

/L =

1 m

g/dL

bili

rubi

n) in

man

ufac

ture

d sa

mpl

es

with

albu

min

conc

entra

tions

of 0

, 10,

and

20 g/

L, re

spec

tivel

y. Th

e med

ian

of m

easu

red

bilir

ubin

conc

entra

tion

is m

arke

d by

the h

oriz

onta

l lin

e in

the c

entra

l box

. The b

oxes

are l

imite

d by

the 2

5th

and

75th

per

cent

ile. Th

e whi

sker

s (┴

) rep

rese

nt th

e low

est a

nd h

ighe

st m

easu

red

bilir

ubin

co

ncen

tratio

ns w

ithin

1.5 i

nter

quar

tile d

istan

ce b

elow

or a

bove

the b

ox. O

utlie

rs (○

) rep

rese

nt b

iliru

bin

conc

entra

tions

bet

wee

n 1.5

and

3 in

terq

uart

ile d

istan

ces b

elow

or a

bove

the b

ox. C

oeffi

cien

ts o

f var

iatio

ns (C

V) a

re sh

own

in fi

g. 1B

and

C as

wel

l as t

he d

iffer

ence

bet

wee

n th

e m

inim

al an

d m

axim

al m

easu

red

bilir

ubin

conc

entra

tion,

whi

ch is

expr

esse

d as

per

cent

age o

f the

wei

ghed

-in b

iliru

bin

conc

entra

tion.

136

Albumin

Table 3 shows mean measured albumin concentrations of samples spiked with bili-rubin ranging from 0 to 500 µmol/L (samples 6 to 20 Table 1). Measured albumin concentrations were ~10% lower than the concentrations of albumin dissolved in the samples. Maximal interlaboratory CV was 6.8%. Unlike the albumin concentra-tion in the bilirubin measurements, bilirubin concentration did not affect albumin measurements. For example, mean (± SD) measured albumin concentration of the samples with an albumin concentration of 10 g/L and 6 different bilirubin concentra-tions (samples 6 to 11 in Table 1) was 8.9 (± 0.6) g/L. For samples with an albumin concentration of 20 g/L (samples 13 to 18 in Table 1), the mean (± SD) albumin concentration was 18.0 (± 0.7) g/L.

Table 3. Measured Albumin Concentrations

Albumin (g/L) Mean ± SD (range) CV (%)

30 27.2 ± 1.0 (26.3 – 29.0) 3.5

25 22.6 ± 0.9 (21.5 – 24.3) 4.0

20* 18 ± 0.7 (17.3 – 19.2) 3.7

15 13.3 ± 0.7 (12.5 – 14.6) 5.3

10* 8.9 ± 0.6 (8.1 – 10.0) 6.8

Data represent results of 10 laboratories for the measured concentrations of EQA-samples 6 to 20 shown in Table 1 with albumin concentrations ranging from 10 to 30 g/L.

Results are expressed as mean SD (ranges) (g/L) with interlaboratory coefficients of variation (CV) per concentration (%).

* Albumin concentrations of 10 and 20 g/L describe mean albumin concentrations of samples with 6 different bilirubin concentrations.

Discussion

This study demonstrates 1) large variability of neonatal bilirubin and albumin con-centration measurements between laboratories of Dutch NICUs potentially affect-ing treatment of jaundiced newborn infants, and 2) lack of standardized devices/ methodology for neonatal bilirubin and albumin measurements on laboratories of Dutch NICUs. Differences in applied laboratory methodology can, theoretically, result in interlaboratory variability. Discrepancies up to 13% of bilirubin measure-ments have been observed in the past between wet chemistry and dry chemistry analyzers using quality control samples.(2) Even greater differences of 30% higher results for wet versus dry bilirubin measurements in human serum samples have

Measurements of neonatal bilirubin and albumin concentrations 137

been demonstrated.(2) In our study, all laboratories used wet chemistry analyzers so the observed and rather large interlaboratory variability is not due to differences in wet versus dry chemistry methodology. Our results are in agreement with data of Vreman, who reported similar interlaboratory variability of bilirubin measurements, applying similar methodology and/or devices.(15)

Interlaboratory variability can be traced, at least in part, to height of the bilirubin concentration and to the use of different matrixes (i.e., bovine versus human serum based), as has been addressed previously.(5–7,13) In 1982 Schreiner et al. studied the bilirubin variability of stabilized liquid quality control sera spiked with different con-centrations of bilirubin. An interlaboratory variability comparable with our findings was found as well as an increased variability in the higher bilirubin concentrations.(13) Matrix effects refer to the difference in behavior of patient samples versus processed (EQA-)samples. In bilirubin analysis, the protein content may affect the molar absorp-tivity of the bilirubin reaction. Frequently used matrices for preparing calibrators are human serum, and human or bovine serum albumin based solutions. The variable and unpredictable underestimation of conjugated bilirubin in bovine serum renders the relationship between the actual bilirubin content and assigned values in calibrators unreliable and dependent on the quality and source of the bovine products. There is no reliable method for accurately measuring the concentration of bilirubin in calibra-tors and controls in bovine sera from commercial sources. The use of these products compromises the accuracy of bilirubin measurements in newborn infants. Therefore human instead of bovine serum should be used for preparing bilirubin calibrators to minimize variability. In 2003, the College of American Pathologists included in the Neonatal Bilirubin Surveys a human serum based sample enriched solely with unconjugated bilirubin.(6,7) Native human serum is the most commutable matrix among different laboratory methods because it mimics the entire serum composition.(6) Yet, use of same matrix does not necessarily imply minimal variability. Recently, the Dutch Foundation for Quality Assessment in Clinical Laboratories initiated a study with two native human serum samples spiked with two different total bilirubin concentrations (30 and 70 µmol/L) to assess standardization of bilirubin measure-ments in the Netherlands among different manufacturers. In analogy to our results, large interlaboratory variability was demonstrated.(4)

Apart from the large interlaboratory variability, underestimation of the bilirubin concentration was observed in albumin free samples. This underestimation of biliru-bin concentration in the albumin free samples is very likely a pre-analytical problem: albumin is essential to dissolve and stabilize bilirubin.

This study illustrates significant interlaboratory differences among bilirubin measurements using aqueous HSA-based samples. Yet, one should be careful with

138

extrapolation of these findings to native neonatal samples as aqueous HSA-based samples behave differently. Moreover, the variability was deduced from the weighed-in calculated bilirubin and albumin concentration, and not by value assignment using an internationally recognized reference method. Notwithstanding these limitations, a large interlaboratory variability of bilirubin and albumin measurements is noticed.

International management guidelines of infants with unconjugated hyperbiliru-binemia are based on total serum bilirubin. However, bilirubin neurotoxicity may be better predicted by the unbound or free bilirubin, i.e., the fraction of bilirubin not bound to plasma proteins (mainly albumin), than by TSB. Unfortunately, free bilirubin measurements are laborious and not routinely available, in contrast to TSB measurements.(16) Inaccurate and, or imprecise bilirubin measurements can potentially result in overtreatment or undertreatment of jaundiced neonates, includ-ing inappropriate blood sampling. Overtreatment exposes the newborn infants to unnecessary risks. Phototherapy, although generally considered safe, may result in retinal damage, skin pathology and an increased insensible water loss.(10) Exchange transfusions are invasive and associated with considerable morbidity and mortality (estimated to occur in 50 and 3 per 1000 procedures, respectively).(1,11) In contrast, underestimation of bilirubin levels may result in undertreatment, possibly contribut-ing to a re-emerge of bilirubin neurotoxicity in jaundiced infants.(8)

Variability in bilirubin and albumin measurements will for obvious reasons nega-tively influence standardizing care for jaundiced newborn infants. Standardized care based on the best available evidence has been demonstrated to improve pediatric and adult patient outcomes.(12) To standardize the management of unconjugated hyperbilirubinemia accurate and precise measurement of bilirubin and albumin con-centrations are inevitably. Monitoring laboratory performance is of key importance herein as has also been internationally recognized.(7)

Recommendations to improve and standardize care of jaundiced newborn infants are, according to the authors of this study, twofold. First, accuracy of bilirubin and albumin measurements needs to be improved and desired quality specifications should be defined. In agreement with international experts, we recommend a total error for neonatal bilirubin of maximal 10% of the reference method.(7) Interestingly, two manufacturers advocated to recalibrate their bilirubin assays for measurement of total bilirubin in the last quarter of 2008. A downscaling of formerly assigned con-centrations to calibrators of -11 to -20% was recommended.(4) Second, commutable external quality control samples with high neonatal bilirubin levels should be used to monitor the required reduction in interlaboratory variability. Although originally recommended in the 20th century, this recommendation is currently still valid for many countries.(15) In the Netherlands, in addition to ‘adult range’ bilirubin and albumin samples, bilirubin and albumin concentrations typically encountered in

Measurements of neonatal bilirubin and albumin concentrations 139

jaundiced neonates, i.e. 200–500 µmoll/L and 15–25 g/L, respectively, are nowadays biweekly monitored by clinical chemistry laboratories involved in care for neonates with unconjugated hyperbilirubinemia.

Conclusions

To comply with international guidelines for the management of neonates with un-conjugated hyperbilirubinemia, exchangeability of bilirubin and albumin measure-ments among laboratories is essential. Our current findings, although in aqueous HSA-samples, apparently point to unacceptably high (interlaboratory) variability of bilirubin and albumin measurements. To analyze and improve the interlaboratory variability a tailor made Quality Assessment Scheme for neonatal samples (human serum based) is available in the Netherlands since January 2010.

Acknowledgements

BARTrial studygroup:Academic Medical Center University of Amsterdam:

Mrs. L. van Toledo - Eppinga, MD, PhD* and J. Fischer, PhD**University Medical Center Maastricht:

A.L.M. Mulder, MD, PhD* and O. Bekers, PhD**Erasmus Medical Center Rotterdam:

P. Govaert, MD, PhD* and R. de Jonge, PhD**Isala Clinics Zwolle:

R.A. van Lingen, MD, PhD* and R.J. Slingerland, PhD**University Medical Center Leiden:

E. Lopriore, MD, PhD* and Mrs. M.C. Cobbaert, PhD**Maxima Medical Center Veldhoven:

J. Buijs, MD* and Prof. H.L. Vader, PhD**University Medical Center Groningen:

Mrs. D.E. van Imhoff, MD and P.H. Dijk, MD, PhD* C.V. Hulzebos, MD, PhD* and L.J. van Pelt, MD, PhD**

University Medical Center St. Radboud Nijmegen: K.D. Liem, MD, PhD* and M.A.M.A. Roelofs - Thijssen, PhD**

University Medical Center Utrecht: Mrs. M.J.N.L. Benders, MD, PhD* and H. Kemperman, PhD**

University Medical Center Amsterdam: Prof. W.P.F. Fetter, MD, PhD* and A.A. Bouman, PhD**

* Pediatrician - neonatologist ** Clinical Chemist

140

References

1. American Academy of Pediatrics. (2004) Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics; 114:297–316.

2. Apperloo JJ, van der Graaf F, Scharnhorst V et al. (2005) Do we measure bilirubin correctly anno 2005? Clin.Chem.Lab Med.43:531–35.

3. Bhutani VK, Maisels MJ, Stark AR et al. (2008) Management of jaundice and prevention of severe neonatal hyperbilirubinemia in infants >or=35 weeks gestation. Neonatology.94:63–67.

4. Cobbaert C, Weykamp C, Hulzebos CV. (2010) Bilirubin Standardization in the Netherlands: Alignment within and between Manufacturers. Clin.Chem.

5. Doumas BT, Eckfeldt JH. (1996) Errors in measurement of total bilirubin: a perennial problem. Clin.Chem.42:845–48.

6. Lo SF, Doumas BT, Ashwood ER. (2004) Performance of bilirubin determinations in US laboratories--revisited. Clin.Chem.50:190–194.

7. Lo SF, Jendrzejczak B, Doumas BT. (2008) Laboratory performance in neonatal bilirubin testing using commutable specimens: a progress report on a College of American Pathologists study. Arch.Pathol.Lab Med.132:1781–85.

8. Maisels MJ. (2009) Neonatal hyperbilirubinemia and kernicterus – Not gone but sometimes forgotten. Early Hum.Dev.

9. Maisels MJ, Bhutani VK, Bogen D et al. (2009) Hyperbilirubinemia in the newborn infant > or =35 weeks’ gestation: an update with clarifications. Pediatrics; 124:1193–98.

10. Maisels MJ, McDonagh AF. (2008) Phototherapy for neonatal jaundice. N.Engl.J.Med.358:920–928.

11. Maisels MJ, Watchko JF. (2003) Treatment of jaundice in low birthweight infants. Arch.Dis.Child Fetal Neonatal Ed; 88:F459-F463.

12. Muething SE. (2005) Improving patient outcomes by standardizing care. J.Pediatr.147:568–70.

13. Schreiner RL, Glick MR. (1982) Interlaboratory bilirubin variability. Pediatrics; 69:277–81.

14. Van Imhoff DE, Dijk PH, Hulzebos CV. (2009) Uniform intervention criteria for jaundice in hyperbilirubinemia in preterm infants. Ned.Tijdschr.Geneeskd.153.

15. Vreman HJ, Verter J, Oh W et al. (1996) Interlaboratory variability of bilirubin measurements. Clin.Chem.42:869–73.

16. Wennberg R, Ahlfors C, Bhutani V et al. (2006) Toward Understanding Kernicterus: A Challenge to Improve the Management of Jaundiced Newborns. Pediatrics; 117:474–85.

567

SD

9

234

8

141

Chapter 9

General discussion

142

Introduction

This thesis shows the results of several studies in preterm infants with unconjugated hyperbilirubinemia. Primarily, evidence on the use of the B/A ratio in the manage-ment of hyperbilirubinemia in preterm infants of less than 32 weeks of gestational age is provided (chapter 3 and 4). Furthermore, recommendations are given on the use of transcutaneous bilirubin levels (chapter 5). The background and synopsis of Dutch consensus-based total serum bilirubin thresholds is described (chapter 6), recommendations on the effective use of phototherapy (chapter 7) and on a quality assessment scheme for laboratory measurements of neonatal bilirubin and albumin (chapter 8) are given. Combining the results of the studies, we propose a set of recom-mendations on the management of hyperbilirubinemia in preterm infants of 35 or less weeks of gestational age (Appendix 1). To establish these studies on hyperbili-rubinemia, the Dutch NICUs collaborated closely and combined their joint efforts in a Dutch Neonatal Research Network (Nederlands Neonatal Research Network, NNRN), in which all 10 Dutch NICUs participate.

Guidelines on the management of hyperbilirubinemia in preterm infants

The management of hyperbilirubinemia in preterm infants is a daily practice for most neonatologists. Since exchange transfusions have become exceptional, phototherapy devices are, as to speak, part of the NICU furniture. As discussed in chapter 6, the TSB thresholds used for preterm infants are far from evidence based. As a result variable guidelines are used among NICUs.(1–14)

The Dutch consensus based treatment thresholds presented in chapter 6 of this thesis are based on TSB level per birth weight category. Lower TSB thresholds are used for infants at risk for developing severe hyperbilirubinemia or bilirubin neurotoxicity. The Dutch thresholds are very similar to the recently published thresholds of Maisels et al. for preterm infants, but lower than the thresholds recently recommended in the UK.(1,15) In contrast to the Dutch thresholds, the American approach and the NICE guideline based their thresholds on gestational age. To be able to compare our thresholds with the ones of Maisels et al. and the NICE, we converted the birth weight categories into gestational age categories based on the mean birth weight and corresponding gestational ages from the Dutch perinatal registry (PRN).(16) Table 1 shows a comparison of these guidelines.(1,15)

General discussion 143

Table 1. Comparison of TSB thresholds for the use of phototherapy in preterm infants

Gestational age NICE 2010 Maisels 2012 Dutch thresholds* Birth weight

<28 130–170 86–103 100 <1000

28 180 103–137 100–150 1000–1250

29 190 103–137 100–150 1000–1250

30 200 137–171 150–190 1250–1500

31 210 137–171 150–190 1250–1500

32 220 171–205 190–210 1500–2000

33 230 171–205 190–210 1500–2000

34 240 205–239 220–240 >2000

35 250 205–239 220–240 >2000

TSB = total serum bilirubin in µmol/L. Gestational age in weeks. Ranges for all the TSB PT thresholds of Maisels are given. The lowest range is used for infants at greater risk to develop severe hyperbili-rubinemia. Risk factors are given in Table 2. The range in TSB PT thresholds of 130 to 170 µmol/L for infants of less than 28 weeks of gestational age of the NICE guideline reflects thresholds for infants of 23 to 28 weeks of gestational age.

* based on the mean birth weight (p50) for each full week of gestational age in the Dutch population of newborns, derived from the PRN database.(16)

The NICE guideline includes recommendations for the management of hyperbiliru-binemia in infants of 23 or more weeks of gestational age, and intended to base these upon studies with the highest available level of evidence. With regard to treatment thresholds, no robust evidence was found in their literature search to determine safe TSB based treatment thresholds for term and preterm infants. Therefore, TSB based PT thresholds were determined according to the formula: phototherapy threshold bilirubin (µmol/L) = (gestational age × 10) − 100. Separate nomograms are designed for infants ranging from 23 to 37 weeks of gestational age. The thresholds rise from 40 µmol/L L at birth to a plateau TSB level after 72 hours.(1)

Since high level of evidence on the management of hyperbilirubinemia in preterm infants is lacking, the recommendations of Maisels et al. did not meet American Academy of Pediatrics (AAP) requirements for guidelines and are thus called an ap-proach. TSB treatment thresholds are lower when compared to the NICE guideline, especially for infants of lower GA.(11,15)

Unlike the recommendations of Maisels et. al and the Dutch, the NICE guideline does not use lower thresholds to start treatment for infants with risk factors. The NICE guideline describes an increased risk for the development of severe hyperbilirubinemia in infants of 38 or less weeks of gestational age, infants with jaundice within the first 24 ours of birth or an increase in severity of clinically apparent jaundice and infants

144

whose parents had the intention to breastfeed exclusively. Although not evidence based, a previous sibling with neonatal jaundice was regarded as an additional risk factor. In addition, a high TSB levels (> 340 µmol/L), rapidly rising TSB levels and clinical features of acute bilirubin encephalopathy increase the risk on kernicterus or other adverse sequelae. These risk factors are included in the NICE guideline. However, the NICE guideline does not specify how to treat these infants at risk dif-ferently.(1) Table 2 shows that the Dutch risk factors are very similar to the ones of Maisels et al..(15) Risk factors may increase the chance to develop hyperbilirubinemia by more bilirubin production or less elimination, or a risk factor may increase the chance that bilirubin enters and harms the brain.(15) Unfortunately, there is only limited and sometimes contradictory evidence on the risk factors that predispose the infants at greater risk to develop severe hyperbilirubinemia or bilirubin neuro-toxicity.(17,18,19,20)

Table 2. Comparison of risk factors for the development of severe hyperbilirubinemia

Maisels et al. 2012 Van Imhoff et al. 2011

• Apnea and bradycardia requiring cardio-respiratory resuscitation in prior 24h

• Mechanical ventilation at the time of blood sampling

• Rapidly rising TSB levels suggesting hemolytic disease

• Blood pH < 7.15 • Sepsis (positive blood culture) in prior 24h • Hypotension requiring pressor treatment

in prior 24h • Lower gestational age • Serum albumin levels < 25 g/L

• Asphyxia (Apgar score < 3 after 5 min) • Hypoxemia (PaO2 < 5.3 kPa > 2 h) in

prior 24 h • Hemolysis with positive Coombs • Blood pH < 7.15 > 1 h in prior 24 h • Sepsis with the use of vasopressors • Meningitis • Intracranial hemorrhage (> grade 2)

h = hours

Conditions that increase the fraction of free bilirubin are associated with low albumin concentrations, or with displacement of bilirubin from albumin (e.g., by sulfonamides or free fatty acids) and conditions which potentially increase the vulnerability to bilirubin neurotoxicity are related to a decreased intactness of the blood-brain barrier (e.g. hyperosmolality, hypercapnia, asphyxia, prematurity, infection, and sepsis) and can result in a net increase of bilirubin uptake in the brain.(21) Oh et al. showed in extreme low birth weight infants that clinical status affected the association between TSB level and neurodevelopmental outcome. High TSB levels in unstable infants were directly associated with an increased risk of death or adverse neurodevelopmental

General discussion 145

outcome, while instable infants this association was absent.(22) Additional clini-cal trials are needed to explore the effect of hyperbilirubinemia and presumed risk factors on neurotoxicity and the neurodevelopmental outcome of preterm infants.

The bilirubin/albumin ratio

The role of the B/A ratio as an additional parameter to the TSB level in the manage-ment of neonatal hyperbilirubinemia has been debated for a long time. On the one hand there is more free bilirubin when albumin levels are low while on the other hand interindividual variations in intrinsic bilirubin-albumin constant and in levels of albumin exists as well as the possibility that TSB binds to alternative binding sites.(23,24) Chapter 3 reviews the literature on the use of the B/A ratio in the manage-ment of hyperbilirubinemia. Abnormal auditory brainstem responses and lower IQ scores at 6 years were associated with high B/A ratios as well as lower albumin levels in infants with postmortem findings of kernicterus. The NICE reviewed the literature as well and included our review as well as two other studies on the use of the B/A ratio in the management of hyperbilirubinemia. These two other studies, an Indian and a Canadian study, showed, in (near) term infants that the B/A ratios and B/A molar ratios correlated significantly to free bilirubin levels (r = 0.74 and 0.75 respectively).(1,25,26) From these data we might conclude that the B/A ratio is valu-able in the management of hyperbilirubinemia. The NICE guideline working group had the opinion that the evidence on the use of the B/A ratio in the management of hyperbilirubinemia was not strong enough to include the B/A ratio in their recom-mendations. Furthermore it was stated that laboratory albumin levels are frequently overestimated and therefore influence the B/A ratio. At the time the NICE guideline was published, they suggested awaiting the results of the BARTrial.(1)

Subsequently to the findings of our review that the B/A ratio may be valuable in the management of hyperbilirubinemia, we conducted a randomized controlled trial on the additional use of the B/A ratio in the management of hyperbilirubinemia in preterm infants. Chapter 4 (BARTrial) shows that the additional use of the B/A ratio is not beneficial in the treatment of hyperbilirubinemia in preterm infants. Infants, who were treated according to the B /A ratio, did not have a better neurodevelopmental outcome as compared to the group treated following the TSB-only thresholds. These results would indicate that it is not valuable to incorporate the B/A ratio thresholds in current guidelines on hyperbilirubinemia in preterm infants. Consequently we did not include the B/A ratio in our set of recommendations (appendix 1) for the management of hyperbilirubinemia. 

However, we found a significantly reduced mortality in the group of infants with a birth weight of more than 1000 grams in the group treated according to the B/A

146

ratio. In contrast, Morris et al. found in their study on aggressive versus prophylactic PT treatment in preterm infants a trend towards increased mortality in infants of less than 750 grams in the aggressive PT group.(27) The reduced mortality in infants of more than 1000 grams in the BARTrial was not associated with less PT, while the increased mortality in the study of Morris et al. may be attributed to more oxidative stress injury to cell membranes due to lower levels of bilirubin which also has an anti oxidant function when moderately increased.(27,28) We cannot explain why infants of more than 1000 grams in the BARTrial showed a reduced mortality. We can only speculate that the reduced mortality may be associated with the additional use of the B/A ratio in preterm infants with hyperbilirubinemia. We compared the BARTrial results on mortality to historical cohorts registered in the Dutch Perinatal Registry (PRN): Mortality in the B/A ratio group of infants weighing more than 1000 gr was also significantly lower compared to the same birth weight group in recent historical cohorts treated according to the novel TSB treatment thresholds.(16) Therefore, ad-ditional clinical trials are needed to explore the role of the B/A ratio and birth weight in preterm infants with hyperbilirubinemia.

Transcutaneous bilirubin levels

Transcutaneous bilirubin measurements are extensively analyzed in term and preterm infants. TcB levels are more accurate in the assessment of jaundice than visual inspec-tion and that the BiliCheck is more accurate than the JM-103.(1) Subsequently, the NICE recommends using TcB levels only in infants of 35 or more weeks of GA, not within the first 24 hours postnatally. They emphasize to measure TSB levels when TcB levels are higher than 250 µmol/L and in all preterm infants of 35 or less weeks of GA.(1) The AAP guideline on hyperbilirubinemia in (near) term infants recom-mends using the TcB levels in the evaluation of hyperbilirubinemia.(11) In their update of the AAP guideline, Maisels et al. underline that the TcB underestimates the TSB level and they describe several methods to correct for this underestimation. One of the methods is using 70% of the TSB threshold as the cut-off level for TcB levels.(29) As described in chapter 5 the NICE guideline also found that the JM-103 PT device consistently underestimates the TSB level of up to 50 µmol/L. None of the guidelines recommends the use of TcB levels in preterm infants, especially not in those receiving PT. In chapter 5 we report that TcB levels show good agreement with TSB levels before, during and after PT in preterm infants. However, it is only useful in reducing the number of blood samples when a cut-off level of TcB plus 50 µmol/L is used at 70% of the TSB threshold. Additional studies are needed to implement the use of TcB measurements in neonatal care as well as to evaluate this implementation.

General discussion 147

Phototherapy

Phototherapy has effectively contributed to the decrease in incidence of kernicterus and BIND.(30) Nevertheless, it is unknown to what extent recommendations on effective PT are adhered to.

We found that the effective use of PT as reflected by the measured irradiance level varies widely among Dutch NICUs with often too low irradiance levels and large distance between PT device and child. As we depict in chapter 7, this can be explained by the lack of awareness on the factors that influence the effective use of PT. PT should be administered to as much naked skin of the infant as possible, and the PT device should be positioned close to the infant (in compliance with the manufacturers’ safety instructions). Irradiance levels should be measured regularly with a radiometer. Unfortunately, there are several types of radiometers used which measure different spectra of light. As a result, studies on irradiance levels of PT are difficult to compare. Therefore, the development of a universal radiometer that can be used with all PT devices to measure the relevant irradiance level is needed.

Like the AAP guideline for (near) term infants, the NICE guideline includes recommendations on the use of PT.(1,11) Also, several studies on the use of PT in preterm infants were reviewed. They found that multiple PT was not more effective than conventional PT and that fiber optic and LED PT shortened the duration of PT compared to conventional PT. Since most nurses and parents prefer fiber optic PT, the NICE guideline recommends the use of fiber optic PT in preterm infants. However, since conventional PT is found to be more effective compared to fiber optic PT in term infants, the effect of PT should be closely monitored.(1) The American approach also includes recommendations with regard to the effective use of PT and emphasizes that sufficient irradiance levels of PT should be provided to prevent an increase in TSB levels. They also stress regular measurement of irradiance levels of the PT devices with an appropriate radiometer. For infants of less than 1000 grams they recommend to use less intensive PT based on the results of the study of Mor-ris et al. on aggressive versus prophylactic PT that found a higher mortality rate in infants of less than 750 grams when they received aggressive PT.(15,27) While both guidelines advocate the effective use of PT, specific irradiance levels of PT devices are not mentioned.

Possible side effects of PT such as allergic diseases, melanocytic nevi, melanoma or skin cancer, patent ductus arteriosus and retinal damage are described.(31) Moreover, the study of Morris et al. shows a higher mortality rate in infants of less than 750 grams treated with aggressive PT.(27) Data from this same trial were used to assess short- and long term outcome of preterm infants treated with different PT devices. Morris et al. showed that LED’s achieved the highest initial decrease in TSB compared to

148

conventional-, spot- and bili blanket devices. Infants treated with conventional PT devices showed worse neurodevelopmental outcome results compared to other PT devices while biliblankets showed slightly better outcome results compared to LEDs. Morris et al. could not find a plausible explanation for these differences in outcome and emphasize that different PT devices need rigorous testing including short- and long term outcome assessments.(28) While PT devices are evolving and improving, concern is growing that very high irradiance levels, may increase the chance of adverse effects. Especially LED PT devices are able to provide very high irradiance levels which seem very effective for the treatment of hyperbilirubinemia, but on the other hand may play a potential risk in the development of adverse effects. Sufficient, but not too much PT seems more beneficial than harmful in the man-agement of hyperbilirubinemia in preterm infants. Again, additional clinical trials including measurement of irradiance levels of PT devices and short- and long term neurodevelopmental outcome assessments are needed to find the optimal doses of PT in preterm infants.

Quality assessment scheme for the measurement of neonatal bilirubin and

albumin

In chapter 8 we describe the need for a quality assessment scheme for laboratory bilirubin and albumin measurements within the neonatal range. To comply with inter-national guidelines for the management of infants with hyperbilirubinemia, reliability and exchangeability of bilirubin and albumin measurements among laboratories are essential. Such quality assessment for neonatal bilirubin and albumin levels did not exist in the Netherlands. We showed in chapter 8 that this absence resulted in high interlaboratory variability in bilirubin and albumin measurements. This could, for obvious reasons, result in over- or under treatment of infants with hyperbilirubine-mia. To analyze and improve the interlaboratory variability, a tailor-made quality assessment scheme for neonatal samples (human serum based) is available in the Netherlands since January 2010.(chapter 8) Now, 64 hospitals in the Netherlands are participating.(oral communication) Although not described in American and UK guidelines on hyperbilirubinemia, quality assessment schemes for neonatal bilirubin do exist in these countries.(32,33)

The Neonatal Research Network

The studies presented in this thesis together propose a set of recommendations on the management of hyperbilirubinemia in preterm infants of less than 35 weeks of gestational age in the Netherlands. To establish this, the Dutch NICUs combined

General discussion 149

their efforts in a Dutch Neonatal Research Network (Nederlands neonatal research network, NNRN).

Simultaneously with the BARTrial, another study was initiated investigating neu-rodevelopmental outcome after neonatal hypoglycemia in all Dutch NICUs (http://www.neonatologiestudies.nl/home/page.asp?page_id=1059). To facilitate the study procedures, collaboration of the participating NICUs (n=10) resulted in the NNRN. The aim of the network was to collaborate in signaling, designing and executing stud-ies in the field of neonatology to accomplish safe and effective neonatal management strategies. All NICUs were represented by one staff neonatologist, research nurses, researchers and other interested parties. The collaboration of NICUs enables the design and execution of larger clinical trials in preterm infants.

Bayley Scales of Infant Development III

Another BARTrial (Chapter 4) network spin-off with major implications in Dutch neonatology is the implementation of the Bayley Scales of Infant Development III. By the time the BARTrial was preparing the follow up of all included infants, the third edition of the Bayley Scale of Infant Development (BSID) was implemented in the United States of America. Although the BSID III seems to underestimate developmental delay, it is thought to better reflect the neurodevelopmental stage of babies and toddlers than earlier editions.(34) The revised test scores the infant at 5 domains: cognitive development, receptive and expressive language development and fine and gross motor development. Furthermore, the test consists of a parent survey about social-emotional development and adaptation of the infant. To keep up with international standards, we decided to use the BSID III instead of the BSID II in the follow up of the BARTrial. The implementation of the second version in The Netherlands took several years. The BSID III was implemented in the Neth-erlands within one year as a result of collaboration between the national follow up workgroup for neonatology (Landelijke Neonatale Follow up, LNF), the NNRN and the BARTrial studygroup.

Future perspectives

Several efforts have been made to improve neonatal care. This resulted in a fast evolu-tion of neonatal care since the introduction of NICUs in 1961. In neonatology, most of the practices are consensus- instead of evidence based due to a lack of evidence on the effectiveness of most practices. The term ‘potentially best practice’ is introduced

150

by Plsek et al. to underline that the process towards ‘evidence based medicine’ in neonatology is ongoing.(35)

In spite of evidence based effective interventions and ‘potentially best practices’, variation in practices still exist among different health care institutions. Variation in care is thought to negatively influence patient outcomes. To accomplish standardiza-tion in care, it is important to compare patient outcomes to different ‘evidence based’ guidelines, recommendations, protocols etc.(36)

Benchmarking is often used as an approach through which variable practices within institutions can be changed into ‘potentially best practice’. Different prac-tices among institutions are explored and the patient or institutional outcomes are compared. Eventually, the ‘potentially best practice’ is identified and adopted by all institutions.(37)

The Vermont Oxford Network is a good example of improving neonatal care through benchmarking by collaboration among NICUs.(38) Since 1995, the Neonatal Intensive Care Collaborative Quality (NIC/Q) Project represented by a neona-tologists, NICU nurses, an administrator, a quality improvement coach and other professionals of the participating NICUs aim to make improvements in the quality of neonatal care. A national database was build and used to evaluate practices and outcomes from which an improvement goal was elected. Based on literature review of preferably randomized controlled trials on the subject, analyses of NICU practices, site visits to participating NICUs and benchmarking visits to superior performing NICUs, a ‘potentially best practice’ is identified to improve neonatal care. One of the NIC/Q projects in 1994 on nosocomial infections showed that in a 3 year period, the infection rate with coagulase-negative staphylococcus decreased from 22% to 16.6% in 6 NICUs. Although this was not a significant change, the decrease in infection rate was significantly larger than observed in 66 comparison NICUs in the same time period.

Table 3. The essential steps towards improvement in neonatal care according to the NIC/Q project.(38,39)

Essential steps

1. Multidisciplinary collaboration within and among hospitals

2. Feedback of information from the network database regarding clinical practice and patient outcome

3. Training and quality improvement methods

4. Site visits to project NICUs

5. Benchmarking visits to superior performers within the network

6. Identification and implementation of ‘potentially best practices’

7. Evaluation of the results

General discussion 151

Horbar et al. conclude that the use of collaborative neonatal quality improvement projects have the potential to improve patient outcome. Table 3 shows the essential steps necessary to improve neonatal care according to the NIC/Q project.(38,39)

Future improvements in the management of hyperbilirubinemia in term and

preterm infants

Concerning the previous outline of processes to improve neonatal care, we have already made some progress in the improvement in the management of preterm in-fants with hyperbilirubinemia. With regard to the uniform treatment thresholds, the recommendations for PT and the neonatal bilirubin and albumin quality assessment scheme the first steps are taken.(39) Future objectives would be to visit the NICUs and the evaluation of the results. By visiting the NICUs, the commitment to the treatment thresholds and recommendations for PT could be verified. If one NICU proves to be better performing with respect to commitment, other NICUs could learn from this NICU how to approach their goal. Benchmarking visits organized by the NNRN may help in this process.

Recommendations for the use of transcutaneous bilirubin measurements

Data on the use of TcB in preterm infants with hyperbilirubinemia described in chapter 5, show that TcB can be used in preterm infants receiving PT. When we used a TcB cut-off level of the measured TcB level plus 50 µmol/L at 70% of the TSB PT threshold we may reduce the number of blood samples with 41%. The next step would be to implement the use of the TcB levels in the management of hyperbilirubinemia in preterm infants in all NICUs.

In the recommendations on the use of TcB measurements using the JM-103, we stress that if manufacturers change the algorithm of the TcB device, this should be openly communicated with clinicians. A change in the algorithm of the TcB device leads to changes in the displayed TcB level and consequently, these changes influence treatment decisions. It is possible that a change in algorithm of the TcB device leads to a change in the method to define the TcB cut-off level. Therefore, it is recommended that every hospital that uses TcB levels in the management of hyperbilirubinemia regularly monitors TcB levels with simultaneous TSB levels.

Recommendations on treatment thresholds

To evaluate the novel treatment thresholds the first step would be to analyze the outcome of patients before and after the implementation of the treatment thresholds. In the Netherlands, perinatal data including treatment for hyperbilirubinemia of all preterm infants of 32 or less weeks of GA admitted to NICUs is collected in the Dutch

152

PRN database.(16) Unfortunately, peak TSB levels, duration of PT and number of PT devices are not registered yet. Recently, neurodevelopmental outcome of al preterm infants of less than 30 weeks of GA or less than 1000 grams is registered in the PRN database. The following items are registered until the child is 1.5 years old: Bayley Scales of Infant Development scores, Childhood behavior Checklist, neurological examination, The Gross Motor Function Classification System (GMFCS) score, visual and auditory examination, general health, length, weight, head circumference, social economic status, language and Automated Auditory Brainstem Response (ALGO or AABR) and if necessary, more advanced auditory brainstem response (ABR) results. Comparison of preterm infants treated for hyperbilirubinemia before and after the implementation of the PT recommendations and the quality assessment scheme with regard to TSB levels, duration of PT and outcome parameters may give us insight in effects of these implementations. Preliminary data on the influence of novel treatment thresholds on the prevalence of hearing deficit exist. It appeared that novel treatment thresholds resulted in a significant decrease in mean and peak TSB levels and a marked reduction in the prevalence of hearing deficit In a retrospective pilot study of preterm infants. This retrospective pilot study needs to be continued in a larger number of patients, preferably by using data from the PRN database.(40)

Recommendations for the effective use of PT

Evaluation of patient outcome before and after the implementation of the recom-mendations for the effective use of PT is difficult because the duration of PT was not registered before the recommendations were introduced and new PT devices (LEDs) could have been introduced at NICUs. Evaluation of the commitment to the recom-mendations is probably the best we can do for now. Special attention should be given to the regular measurement of the irradiance levels of PT and the distance between the device and the infant. To the best of our knowledge, there is still no universal radiometer that measures the irradiance level of all PT devices in the clinical relevant treatment spectrum. The development of such a radiometer is one of the main goals in improving the effectiveness of PT treatment.(1,41,42) Training in the knowledge of effective PT treatment of all persons concerned with the care of preterm infants at the NICU is another basic need. The recommendations on the effective use of PT in NICUs are easily applicable in all pediatric healthcare facilities and therefore should be available for all healthcare facilities concerning newborn infants.

General discussion 153

Quality assessment scheme for the measurement of neonatal bilirubin and

albumin

The quality assessment scheme (QAS) for the measurement of neonatal bilirubin and albumin is in use since January 2010. The next step would be to evaluate the results of this QAS in the participating hospitals to assess whether the implementa-tion of the QAS has resulted in the desired reduction of variability of the bilirubin and albumin measurements.

154

References

1. National Institute for Health and Clinical Excellence. National Institute for Health and Clinical Excellence. Neonatal Jaundice (Clinical guideline 98). www.nice.org.uk/CG98. 2010.

2. Pearlman MA, Gartner LM, Lee K, Eidelman AI, Morecki R, Horoupian DS. The association of kernicterus with bacterial infection in the newborn. Pediatrics 1980 01;65(1):26–29.

3. Pearlman MA, Gartner LM, Lee K, Morecki R, Horoupian DS. Absence of kernicterus in low-birth weight infants from 1971 through 1976: comparison with findings in 1966 and 1967. Pediatrics 1978 10;62(4):460–464.

4. Maisels MJ. Clinical studies of the sequelae of hyperbilirubinemia. In: Levine RL, editor. Hyperbilirubinemia in the newborn, Report of the 85th Ross Conference on Pediatric Research Columbus, OH: Ross Laboratories; 1983. p. 26–38.

5. Watchko JF, Oski FA. Kernicterus in preterm newborns: past, present, and future. Pediatrics 1992 11;90(5):707–715.

6. Watchko JF, Claassen D. Kernicterus in premature infants: current prevalence and relationship to NICHD Phototherapy Study exchange criteria. Pediatrics 1994 06;93(6):996–999.

7. Hansen TW. Therapeutic approaches to neonatal jaundice: an international survey. Clin Pediatr (Phila) 1996 06;35(6):309–316.

8. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003 11;88(6):F455-F458.

9. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal Neonatal Ed 2003 11;88(6):F459-F463.

10. Rennie JM, Seghal A, De A, Kendall GS, Cole TJ. Range of UK practice regarding thresholds for phototherapy and exchange transfusion in neonatal hyperbilirubinaemia. Arch Dis Child Fetal Neonatal Ed 2008 11/10.

11. American Academy oP. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004 07;114(1):297–316.

12. Bratlid D, Nakstad B, Hansen TW. National guidelines for treatment of jaundice in the newborn. Acta Paediatr 2011 Apr;100(4):499–505.

13. Dani C, Poggi C, Barp J, Romagnoli C, Buonocore G. Current Italian practices regarding the management of hyperbilirubinaemia in preterm infants. Acta Paediatr 2011 May;100(5):666–669.

14. van Imhoff DE, Dijk PH, Hulzebos CV, on behalf of the BARTrial studygroup of the Netherlands Neonatal Research Network. Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a national guideline. Early Hum Dev 2011 Aug;87(8):521–525.

15. Maisels MJ, Watchko JF, Bhutani VK, Stevenson DK. An approach to the management of hyperbilirubinemia in the preterm infant less than 35 weeks of gestation. J Perinatol 2012 Sep;32(9):660–664.

16. The Netherlands Perinatal Registry. www.perinatreg.nl. 2009.

General discussion 155

17. Van de Bor M, Ens-Dokkum M, Schreuder AM, Veen S, Brand R, Verloove-Vanhorick SP. Hyperbilirubinemia in low birth weight infants and outcome at 5 years of age. Pediatrics 1992 03;89(3):359–364.

18. Van de Bor M, van Zeben-van der Aa TM, Verloove-Vanhorick SP, Brand R, Ruys JH. Hyperbilirubinemia in preterm infants and neurodevelopmental outcome at 2 years of age: results of a national collaborative survey. Pediatrics 1989 06;83(6):915–920.

19. Graziani LJ, Mitchell DG, Kornhauser M, Pidcock FS, Merton DA, Stanley C, et al. Neurodevelopment of preterm infants: neonatal neurosonographic and serum bilirubin studies. Pediatrics 1992 Feb;89(2):229–234.

20. O’Shea TM, Dillard RG, Klinepeter KL, Goldstein DJ. Serum bilirubin levels, intracranial hemorrhage, and the risk of developmental problems in very low birth weight neonates. Pediatrics 1992 Dec;90(6):888–892.

21. Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 2005 01;25(1):54–59.

22. Oh W, Stevenson DK, Tyson JE, Morris BH, Ahlfors CE, Bender GJ, et al. Influence of clinical status on the association between plasma total and unbound bilirubin and death or adverse neurodevelopmental outcomes in extremely low birth weight infants. Acta Paediatr 2010 May;99(5):673–678.

23. Odell GB, Cukier JO, Ostrea EM,Jr, Maglalang AC, Poland RL. The influence of fatty acids on the binding of bilirubin to albumin. J Lab Clin Med 1977 Feb;89(2):295–307.

24. Robertson A, Karp W, Brodersen R. Bilirubin displacing effect of drugs used in neonatology. Acta Paediatr Scand 1991 12;80(12):1119–1127.

25. Malik GK, Goel GK, Vishwanathan PN, Misra PK, Sharma B. Free and erythrocyte-bound bilirubin in neonatal jaundice. Acta Paediatr Scand 1986 Jul;75(4):545–549.

26. Chan G, Ilkiw R, Schiff D. Clinical relevance of the plasma reserve albumin binding capacity for bilirubin (RABC) and “free” bilirubin concentration. Clin Biochem 1980 Dec;13(6):292–294.

27. Morris BH, Oh W, Tyson JE, Stevenson DK, Phelps DL, O’Shea TM, et al. Aggressive vs. conservative phototherapy for infants with extremely low birth weight. N Engl J Med 2008 10/30;359(18):1885–1896.

28. Morris BH, Tyson JE, Stevenson DK, Oh W, Phelps DL, O’Shea TM, et al. Efficacy of phototherapy devices and outcomes among extremely low birth weight infants: multi-center observational study. J Perinatol 2012 Apr 12.

29. Maisels MJ, Bhutani VK, Bogen D, Newman TB, Stark AR, Watchko JF. Hyperbilirubinemia in the newborn infant > or =35 weeks’ gestation: an update with clarifications. Pediatrics 2009 10;124(4):1193–1198.

30. Maisels MJ, McDonagh AF. Phototherapy for neonatal jaundice. N Engl J Med 2008 02/28;358(9):920–928.

31. Xiong T, Qu Y, Cambier S, Mu D. The side effects of phototherapy for neonatal jaundice: what do we know? What should we do? Eur J Pediatr 2011 Oct;170(10):1247–1255.

156

32. Lo SF, Jendrzejczak B, Doumas BT. Laboratory performance in neonatal bilirubin testing using commutable specimens: a progress report on a College of American Pathologists study. Arch Pathol Lab Med 2008 11;132(11):1781–1785.

33. Kirk JM. Neonatal jaundice: a critical review of the role and practice of bilirubin analysis. Ann Clin Biochem 2008 Sep;45(Pt 5):452–462.

34. Anderson PJ, De Luca CR, Hutchinson E, Roberts G, Doyle LW, Victorian Infant Collaborative Group. Underestimation of developmental delay by the new Bayley-III Scale. Arch Pediatr Adolesc Med 2010 Apr;164(4):352–356.

35. Plsek PE. Quality improvement methods in clinical medicine. Pediatrics 1999 Jan;103(1 Suppl E):203–214.

36. Muething SE. Improving patient outcomes by standardizing care. J Pediatr 2005 11;147(5):568–570.

37. Walsh MC. Benchmarking techniques to improve neonatal care: uses and abuses. Clin Perinatol 2003 Jun;30(2):343–50, x.

38. Horbar JD, Rogowski J, Plsek PE, Delmore P, Edwards WH, Hocker J, et al. Collaborative quality improvement for neonatal intensive care. NIC/Q Project Investigators of the Vermont Oxford Network. Pediatrics 2001 Jan;107(1):14–22.

39. Soll RF. Evaluating the medical evidence for quality improvement. Clin Perinatol 2010 Mar;37(1):11–28.

40. Hulzebos CV, van Dommelen P, Verkerk PH, Dijk PH, van Straaten HLM. Evaluation of treatment thresholds for unconjugated hyperbilirubinemia on hearing loss in preterm infants. 2012.

41. Vreman HJ. Personal communication. 2012.

42. Bhutani VK, Committee on Fetus and Newborn, American Academy of Pediatrics. Phototherapy to prevent severe neonatal hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2011 Oct;128(4):e1046–52.

General discussion 157

Appendix 1.

Recommendations for the management of hyperbilirubinemia in preterm infants of less than 35 weeks of gestational age

1. Select the appropriate thresholds based on birth weight of the infant* (Figure a)2. Regularly measure the TSB level**3. After 48 postnatal hours use the TSB or TcB level → if the TcB level is used, see

Table b for recommendations4. Mark the risk status in the nomogram (Table a)5. Phototherapy:

a. Start PT if the PT threshold is reachedb. Stop PT if TSB is 50 µmol/L lower than the threshold

→ see Table c for the recommendations on the effective use of PT6. Consider an exchange transfusion if the ET threshold is reached despite intensive PT

* more detailed nomograms can be found in chapter 6 and on www.babyzietgeel.nl** participation of the laboratory in a regular quality assessment scheme for neonatal bilirubin and

albumin measurements is highly recommended.

158

Figure a. TSB based treatment thresholds for preterm infants of less than 35 weeks of gestational age TSB = Total Serum Bilirubin in µmol/L, PT = phototherapy, WT = exchange transfusion Use the lowest TSB threshold for PT and ET if risk factors to develop severe hyperbilirubinemia or bilirubin neurotoxicity are present. More detailed nomograms for all birth weight categories can be found in chapter 6 and on www.babyzietgeel.nl

General discussion 159

Table a. Risk factors for the development of severe hyperbilirubinemia or bilirubin neurotoxicity

Risk factors

• Asphyxia: Apgar score < 3 after 5 minutes • Hypoxemia: PaO2 < 5.3 kPa > 2 h (in prior 24h) • Acidosis: Blood pH < 7.15 > 1 h (in prior 24h) • Hemolysis with positive Coombs • Clinical or neurological deterioration (sepsis with the use of vasopressors, meningitis,

intracranial hemorrhage > grade 2)

h= hours

Table b. Recommendations for the use of JM-103 TcB measurements in preterm infants with and without PT

Recommendations TcB

• Measure TcB levels under the diaper on the hipbone of the infant• TcB cut-off level: Add 50 µmol/L to the measured TcB level at 70% of the PT threshold• Ensure regular calibration of the TcB device according to the recommendations of the

manufacturer • Regularly perform paired TSB and TcB levels to check the accuracy of the TcB device• Teach NICU nurses and attending physicians how to perform and interpret TcB mea-

surements• Measure the TSB level:

–If the TcB level plus 50 µmol/L exceeds 70% of the TSB PT threshold –If there is clinical concern on severe hyperbilirubinemia –If you don’t trust the result of that specific transcutaneous measurement

TSB = total serum bilirubin, TcB = transcutaneous bilirubin, PT = phototherapy, NICU = neonatal intensive care unit, 17.1 µmol/L = 1 mg/dL bilirubin

Table c. Recommendations for the effective use of phototherapy

Recommendations PT

• Educate all health care professionals who buy, apply and maintain PT devices about the factors that affect the efficacy of PT.

• Measure the irradiance level of your PT devices regularly.• The minimal recommended irradiance level is 8 to 10 µW/cm2/nm.• Replace lamps that do not deliver.• Match PT devices to incubators in use and place the PT device as close to the infant as

is safe and possible.• Follow the safety instructions of the manufacturer of the PT device to avoid heat and

burns (especially with halogen lights)!• Illuminate as much naked skin of the infant as possible, making sure optimal body

temperature is maintained.• Cover the eyes of the newborn infant.

PT = phototherapy

567

9

DS

234

8

161

Summary

The management of hyperbilirubinemia in preterm infants is a daily practice for most neonatologists. An excess amount of unconjugated bilirubin in the blood results in yellow discoloration of the skin, i.e. jaundice, which can be observed in neonatal hyperbilirubinemia. Hyperbilirubinemia is very common in preterm infants and can potentially harm the central nervous system. The clinical spectrum of signs of bilirubin neurotoxicity relates to the bilirubin-induced damage to specific brain areas. Although acute kernicterus is an unambiguous clinical disorder in severely jaundiced newborn infants with the possibility of permanent sequelae, subtle forms of bilirubin-induced neurological dysfunction (BIND) have been described more recently. BIND can present with auditory dysfunction and/or mild neurologic ab-normalities such as mild impairment in neurologic and/or cognitive performance. The risk of kernicterus and BIND may be in part determined by the concentration of Total Serum Bilirubin (TSB), which in neonates consists almost exclusively of unconjugated bilirubin (UCB), but is primarily determined by the concentration of non-albumin bound free bilirubin (Bf). Bf can easily pass the blood-brain barrier, and may better reflect the bilirubin load distributed in the brain. Unfortunately, measurement of Bf is clinically not available and treatment thresholds are based on TSB levels. However, the bilirubin/albumin (B/A) ratio has been suggested as an approximate parameter of Bf: in case of low albumin levels more Bf is available to cross the blood brain barrier. Current management guidelines for near term infants (> 35 weeks GA) consider low albumin levels as risk factors for the development of bilirubin neurotoxicity. Although the B/A ratio may be valuable in the risk assess-ment of neonates with hyperbilirubinemia, it is not incorporated into guidelines for the management of hyperbilirubinemia.

Preterm infants are more prone to neurological impairment as well as bilirubin neurotoxicity than their term counterparts. In this thesis variable studies are pre-sented with the aim to improve the management of neonatal hyperbilirubinemia in preterm infants of less than 32 weeks of gestational age. The basis of the thesis was this lack of evidence on the use of the B/A ratio in relation to neurodevelop-mental outcome in preterm infants. In chapter 3 we reviewed the literature on the use of the B/A ratio in addition to TSB in the management of hyperbilirubinemia in preterm infants. Abnormal Auditory Brainstem Responses and lower IQ scores

162

at 6 years were associated with high B/A ratios as well as lower albumin levels in infants with postmortem findings of kernicterus. However, considering the lack of prospective clinical trials supporting the clinical benefit regarding B/A ratio and outcome, a randomized controlled trial investigating the concurrent use of the B/A ratio and the TSB in the management of preterm infants with hyperbilirubinemia (BARTrial, Chapter 4) was conducted. In chapter 4 we aimed to determine the ef-fects of the B/A ratio in addition to TSB in the management of preterm infants with hyperbilirubinemia on their neurodevelopmental outcome at 18 to 24 months after the expected date of delivery. We found no significant difference in the rate of the composite motor score at 18 to 24 months of age between the infants assigned to the B/A ratio compared to the TSB group. These results indicate that it is not valuable to incorporate the B/A ratio thresholds in current guidelines on hyperbilirubinemia in preterm infants. However, we found a significantly reduced mortality in the group of infants with a birth weight of more than 1000 grams in the group treated according to the B/A ratio which we could not explain from a pathofysiological point of view. Therefore, additional clinical trials are needed to explore the role of the B/A ratio and birth weight in preterm infants with hyperbilirubinemia.

Non-invasive transcutaneous bilirubin measurements are extensively analyzed in term and preterm infants. There is a paucity of evidence based data on the use of TcB measurements in preterm infants, especially when receiving phototherapy. This was the basis of chapter 5 of this thesis in which we analyzed the relation of TcB and TSB measurements before, during and after phototherapy. Furthermore we determined effects of specific TcB cut-off levels regarding the reduction in the need for blood samples without missing severe hyperbilirubinemia. We found that TcB levels mea-sured on the covered hipbone can be used in preterm infants receiving phototherapy. This is only useful in safely reducing the number of blood samples without missing high TSB levels when a cut-off level of the calculated TcB plus 50 µmol/L is used at 70% of the TSB threshold. Additional studies are needed to implement the use of TcB measurements in neonatal care as well as to evaluate this implementation.

Due to lack of evidence based TSB thresholds for preterm infants with hyper-bilirubinemia, variation in applied TSB treatment thresholds exists. Treatment thresholds for preterm infants with hyperbilirubinemia had to be uniform in order to conduct the BARTrial. Chapter 6 explores the applied TSB treatment thresholds in the Dutch neonatal intensive care units (NICUs). Large variation in applied TSB threshold existed in the NICUs. Novel, consensus-based TSB treatment thresholds for hyperbilirubinemia in preterm infants of less than 35 weeks of GA were developed.

Next to implementation of the consensus-based TSB treatment thresholds, clinical practice conditions of phototherapy were assessed and irradiance levels of

Summary 163

phototherapy devices were measured in all NICUs. Chapter 7 presents our findings that PT devices in the Dutch NICUs show considerable variability with often too low irradiance levels. This can be explained by the fact that phototherapy devices are placed too far away from the infant and the lack of awareness among healthcare workers on the other factors that influence the effective use of PT. Furthermore, ir-radiance levels should be measured regularly with a radiometer.

To further standardize the care of hyperbilirubinemia in preterm infants, reliabil-ity and exchangeability of bilirubin and albumin measurements among laboratories was essential. Quality assessment for neonatal bilirubin and albumin levels did not exist in the Netherlands. In chapter 8 we analyzed the laboratory measurement of bilirubin and albumin levels in the neonatal range of the 10 NICUs and found high interlaboratory variability. This could, for obvious reasons, result in over- or under treatment of infants with hyperbilirubinemia. To analyze and improve the interlabo-ratory variability a tailor made Quality Assessment Scheme for neonatal samples is available in the Netherlands now.

The studies presented in this thesis together propose a set of recommendations on the management of hyperbilirubinemia in preterm infants of less than 35 weeks of gestational age in the Netherlands. To establish this, the Dutch NICUs combined their efforts in a Dutch Neonatal Research Network (Nederlands neonatal research network, NNRN). Future objectives would be to organize benchmark visits to the NICUs to evaluate the results of the recommendations for the management of hyperbilirubinemia in preterm infants. Furthermore, new studies in the field of hyperbilirubinemia should be organized by the NNRN to further improve the care and neurodevelopmental outcome of preterm infants.

164

1234567

9

DS

8

165

Nederlandse samenvatting

De studies in dit proefschrift geven een aantal aanbevelingen weer ter verbeter-ing van de diagnostiek en behandeling van hyperbilirubinemie bij veel te vroeg (prematuur) geboren baby’s in Nederland geboren na een zwangerschapsduur van minder dan 35 weken.

Bilirubine is een afbraakproduct van rode bloedcellen. Na de geboorte worden de rode bloedcellen van baby’s vervangen door volwassen rode bloedcellen. Er gaan dus veel rode bloedcellen te gronde waardoor er ook veel bilirubine vrijkomt. De lever zorgt ervoor dat bilirubine wordt omgezet in een vorm die gemakkelijk via de ontlas ting kan worden uitgescheiden. Bij pasgeboren baby’s is de lever vaak nog niet in staat om de grote hoeveelheid bilirubine op te ruimen. Een overschot aan bilirubine in het bloed (hyperbilirubinemie) zorgt ervoor dat pasgeboren baby’s een gele verkleuring van de huid kunnen krijgen. Dit noemen we ‘geelzien’ of icterus. Hyperbilirubinemie komt erg veel voor bij prematuur geboren baby’s en daardoor is de behandeling van hyperbilirubinemie dagelijkse kost voor de meeste kinderartsen-neonatologen. De behandeling van hyperbilirubinemie bestaat uit lichttherapie (fototherapie) of soms een wisseltransfusie, waarbij het bloed waarin heel veel bilirubine zit, vervangen wordt door ‘nomaal’ donorbloed.

Ernstige hyperbilirubinemie kan schadelijk zijn voor de hersenen (bilirubine neurotoxiciteit). Het klinische beeld hiervan wordt veroorzaakt door bilirubine-geïnduceerde schade in specifieke hersengebieden, waaronder de zogenaamde basale kernen (die ook geel verkleuren, vandaar de naam ‘kernicterus’). Acute kernicterus is een onmiskenbaar klinisch beeld bij baby’s met een heel ernstige hyperbilirubinemie. Soms kan dit leiden tot permanent letsel in de vorm van een spastische verlamming. Voorbeelden van meer subtiele vormen van bilirubine-geïnduceerde neurologische schade zijn gehoorproblemen en milde neurologische afwijkingen zoals een motorische en/of verstandelijke ontwikkelingsachterstand.

Het risico op het ontwikkelen van hersenschade door bilirubine wordt voor een deel bepaald door het totale serum bilirubine (TSB) gehalte. Bij pasgeboren baby’s bestaat het TSB voornamelijk uit bilirubine gebonden aan het eiwit albumine.

166

Een klein deel wordt gevormd door de concentratie vrije bilirubine (Bfree of Bf) dat niet gebonden is aan albumine. Niet het TSB, maar wel het Bf kan gemakkelijk de bloed-hersenbarrière passeren en in de hersenen doordringen. Helaas is het nog niet mogelijk om Bf op een gemakkelijke manier te meten in het bloed. Een andere maat die we zouden kunnen gebruiken om de concentratie Bf te benaderen is de verhouding tussen het TSB en het albumine, de bilirubine/albumine (B/A) ratio. Als er weinig albumine beschikbaar is voor bilirubine om aan te binden, dan is er meer Bf beschikbaar om de bloed-hersenbarrière te passeren.

In overeenstemming met bovenstaande, staan er strengere behandelgrenzen in de richtlijnen voor baby’s geboren na 35 weken zwangerschapsduur met hyperbili-rubinemie die een laag albumine gehalte hebben. Het is grotendeels nog onduidelijk of de B/A ratio een waardevolle toevoeging zou kunnen zijn in de risico-inschatting van bilirubine neurotoxiciteit voor prematuur geboren kinderen. Daarom is de B/A ratio nog niet opgenomen in richtlijnen voor de behandeling van hyperbilirubinemie bij prematuur geboren baby’s.

Prematuur geboren baby’s zijn gevoeliger voor neurologische schade en door bili-rubine veroorzaakte hersenschade dan voldragen pasgeboren baby’s. In hoofdstuk 2 geven we achtergrondinformatie van bilirubine neurotoxiciteit, het klinische beeld en de diagnostische mogelijkheden om de ernst van de hyperbilirubinemie te evalueren.

Het gebrek aan wetenschappelijk bewijs voor de relatie tussen het gebruik van de B/A ratio bij de behandeling van prematuur geboren baby’s met hyperbilirubinemie en de neurologische uitkomst op 2-jarige leeftijd van deze baby’s was de basis voor dit proefschrift. In hoofdstuk 3 hebben we de beschikbare literatuur over het gebruik van de B/A ratio bij de behandeling van prematuur geboren baby’s met hyperbilirubine-mie en het effect daarop op de neurologische uitkomst van deze baby’s geanalyseerd. Een abnormale gehoortest en een lager IQ op 6 jarige leeftijd waren geassocieerd met hoge B/A ratio’s. Verder werden er lage albumine concentraties gevonden bij overleden baby’s met kenmerken van kernicterus. Helaas waren er geen klinische studies bekend over het betrekken van de B/A ratio bij de behandeling van prematuur geboren baby’s met hyperbilirubinemie en de neurologische uitkomst. Dit was voor ons de reden om een landelijke gerandomiseerde studie op te zetten waarin we het effect van de B/A ratio naast het gebruik van het TSB op de neurologische uitkomst geanalyseerd hebben bij baby’s met hyperbilirubinemie (de Bilirubine Albumine Ratio Trial, afgekort de BARTrial, hoofdstuk 4).

We hebben geen verschillen gevonden in de motorische en verstandelijke ontwik-keling van kinderen op de leeftijd van 18 tot 24 maanden, die na de geboorte ook behandeld waren volgens de B/A ratio vergeleken met kinderen die behandeld waren op basis van de TSB concentratie alleen. Op basis van deze resultaten is het

Nederlandse samenvatting 167

niet zinvol om de B/A ratio op te nemen in de richtlijnen voor de behandeling van hyperbilirubinemie bij prematuur geboren baby’s. We hebben in deze studie ook gevonden dat er minder kinderen waren overleden in de groep baby’s met een ge-boortegewicht hoger dan 1000 gram die behandeld waren op basis van de B/A ratio ten opzichte van de baby’s die behandeld waren op basis van de TSB concentratie. Dit konden we niet goed verklaren en kan berusten op toeval. Er zijn vervolgstudies nodig om de rol van de B/A ratio en het geboortegewicht bij premature kinderen met hyperbilirubinemie nader te analyseren.

Met behulp van een transcutane bilirubine (TcB) meter is het mogelijk om op een pijnloze manier, namelijk via de huid, de TSB concentratie in het bloed te berekenen. Op die manier zouden deze baby’s minder vaak geprikt kunnen worden. Er is nog niet zoveel onderzoek verricht naar het gebruik van TcB metingen bij prematuur geboren baby’s. Vooral als ze ook behandeld worden voor hyperbilirubinemie met behulp van fototherapie, omdat dit mogelijk de bepaling via de huid verstoord (de huid wordt namelijk minder geel tijdens de behandeling). In hoofdstuk 5 hebben we eerst de relatie tussen TcB en TSB metingen voor, tijdens en na het gebruik van fototherapie bepaald. Daarna hebben we het aantal bloedafnames berekend die we zouden kunnen besparen door TcB waarden te gebruiken. Het bleek dat TcB metingen op de met een luier bedekte heup van prematuur geboren baby’s vergelijkbaar waren zowel voor, tijdens als na fototherapie. Als we TcB waarden willen gebruiken om het aantal bloedafnames te verminderen, is het van belang er rekening mee te houden dat de kans om hoge TSB waarden te missen minimaal is. We hebben gevonden dat we ongeveer 40% van de bloedafnames konden verminderen met een kleine kans op het missen van een hoge TSB waarde als er 50 µmol/L bij de gemeten TcB waarde wordt opgeteld en de behandelgrens met 30% verlaagd wordt.

In hoofdstuk 6 hebben we de behandelgrenzen voor hyperbilirubinemie bij pre-matuur geboren baby’s in de Nederlandse neonatale intensive care units (NICU’s) geïnventariseerd. Behandelgrenzen voor prematuur geboren baby’s met hyperbiliru-binemie moesten gelijk worden in de verschillende NICU’s voordat we de BARTrial konden uitvoeren. Na de inventarisatie van de gebruikte richtlijnen in de NICU’s, zijn we met alle NICU’s nieuwe gelijke TSB behandelgrenzen overeengekomen voor prematuur geboren baby’s geboren na een zwangerschapsduur van minder dan 35 weken. Ongewenste verschillen in behandeling tussen verschillende ziekenhuizen worden hierdoor verminderd.

Naast het invoeren van deze gelijke TSB behandelgrenzen, hebben we in alle NICU’s de effectiviteit van fototherapie geanalyseerd. We hebben van alle beschik-bare fototherapielampen de stralingssterkte gemeten. Deze studie beschrijven we in hoofdstuk 7. Er was grote variatie in de gemeten stralingssterkte tussen de beschik-

168

bare lampen en tussen de NICU’s. Ook werd er vaak een te lage stralingssterkte gemeten. Dit kunnen we deels verklaren doordat de fototherapielampen vaak op een te grote afstand van de baby geplaatst worden en mogelijk ook doordat artsen en verpleegkundigen soms onvoldoende kennis hebben van de factoren die de effectiviteit van fototherapie beïnvloeden. In dit hoofdstuk hebben we een aantal aanbevelingen gedaan voor het effectief geven van fototherapie en om de kennis daarover bij zorgverleners te vergroten.

Om de zorg voor prematuur geboren baby’s met hyperbilirubinemie nog meer te standaardiseren was het nodig om de bilirubine en albumine metingen in de diverse laboratoria te standaardiseren. In Nederland wordt voor veel laboratoriumbepalingen regelmatig kwaliteitsserum rondgestuurd om deze te ijken. Voor bilirubine en albu-mine concentraties die voorkomen bij pasgeboren baby’s en heel anders zijn dan de waarden bij volwassenen, bestond nog geen kwaliteitsserum. In hoofdstuk 8 hebben we de laboratoria van de NICU’s verschillende neonatale bilirubine en albumine concentraties laten meten. Hieruit bleek dat er een grote variatie in de meetuit-komsten tussen de deelnemende laboratoria bestond. Hierdoor werden baby’s met hyperbilirubinemie in de verschillende NICU’s niet op dezelfde manier behandeld. Om de verschillen tussen de laboratoria te analyseren en te verbeteren bestaat er nu een Nederlands kwaliteitsserum met concentraties bilirubine en albumine zoals die voorkomen bij pasgeboren baby’s.

De door dit proefschrift tot stand gekomen standaardisering van de zorg voor hyper-bilirubinemie bij prematuur geboren kinderen was mogelijk door de samenwerking van de Nederlandse NICU’s in het Nederlandse Neonatale Research Netwerk (NNRN). In de toekomst is het de bedoeling om deze aanbevelingen voor de diagnos tiek en be-handeling van prematuur geboren baby’s met hyperbilirubinemie te blijven evalueren.

1234567

9SD

8

169

Dankwoord

Zonder hulp van velen was dit proefschrift niet tot stand gekomen! Een aantal van jullie wil ik graag noemen.

Allereerst mijn copromotoren dr. P.H. Dijk en dr. C.V. Hulzebos. Toen jullie mij benaderden voor de BARTrial met daaruit voorvloeiend dit promotietraject heb ik niet lang getwijfeld. De praktische uitvoer van het onderzoek en het contact met de patiënten trokken me meteen aan. Het verzamelen, analyseren en opschrijven van data was nieuw voor me en een grote uitdaging. Ik heb ontzettend veel van jullie geleerd in de afgelopen jaren. Dit heeft me als mens en als dokter gevormd. Ik be-wonder jullie tomeloze inzet, enthousiasme en interesse in het doen van onderzoek.

Beste Peter, de uren die we samen sleutelden aan teksten, tabellen en figuren waren zeer leerzaam, enthousiasmerend en motiverend om stukken tot een beter, duidelijker en mooier geheel te maken. Ik heb veel van je geleerd. Dank je wel voor je vertrouwen!

Beste Chris, zonder jouw kritische blik was ik nooit zover gekomen met mijn proefschrift. Elke bespreking weer wist je het stuk naar een hoger niveau te tillen door een andere kant van het onderzoek te belichten. Dank je wel!

Dan mijn promotor prof. dr. A.F. Bos. Beste Arie, toen ik net begon met mijn promotieonderzoek vertelde je dat onderzoek doen vooral uit jezelf moet komen. Deze wijze woorden heb ik zeker ter harte genomen, en hebben me gebracht bij het resultaat dat ik nooit had durven dromen: mijn eigen boekje! Dank je wel voor je steun, kritische blik en neutrale visie waar nodig in mijn promotietraject.

Ik heb grote bewondering voor de ouders en patiënten die mee hebben gedaan met de BARTrial. Het was al moeilijk genoeg dat een kind onverwacht zo vroeg ter wereld kwam, laat staan dat je dan als ouder ook nog na moet denken of je mee wilt doen aan wetenschappelijk onderzoek. Heel hartelijk dank!

De leden van de beoordelingscommissie prof. dr. W.P. Fetter, prof. dr. F.J. Walther en prof. dr. H.J. Verkade wil ik hartelijk danken voor het kritisch doorlezen van mijn proefschrift.

170

De BARTrial studiegroep wil ik danken voor de fijne samenwerking: Leiden UMC: E.Lopriore, Amsterdam MC: L. van Toledo-Eppinga en D.H.G.M. Nuyte mans, UMC Utrecht: M.J.N. Benders en K.K.M. Korbeeck, Isala Zwolle: R. A. van Lingen en L.J.M. Groot-Jebbink, UMC St. Radboud Nijmegen: K.D. Liem en P. Mansvelt, Maxima MC Veldhoven: J. Buijs, Erasmus MC Rotterdam: P. Govaert en I. van Vliet, UMC Maastricht: A.L.M. Mulder en C. Wolfs, VUMC Amsterdam: W.P.F. Fetter en A.R.C. Laarman, Toronto: M. Offringa. Voor de follow up van alle kin-deren: A.G. van Wassenaer-Leemhuis, S.A.J. Ruiter (en alle studenten die hebben geholpen met testen) en K.N.J.A. van Braeckel. En voor de statistiek: P.F.M. Krabbe en E.H. Quik. Jorien, dank je wel voor de ASQ’s. Christa, dank je wel voor de tips voor het neurologisch onderzoeken van jonge kinderen. Henk Breukelman, Theo en Marissa van het lab, dank jullie wel voor de hulp bij de opslag van de monsters! En natuurlijk alle andere neonatologen, arts-assistenten, verpleegkundigen, secre-taresses en anderen van het UMCG en de andere centra die zich voor de BARTrial en de andere studies hebben ingezet: dank jullie wel!

Miriam, dank je wel voor het BART-logo!

Beste Frans Cuperus, dank je wel voor de samenwerking. Gelukkig zijn er, ondanks de soms uitzichtloze uren die we hebben doorgebracht in het lab, een paar mooie stukken tot stand gekomen.

Beste Henk Vreman, hartelijk dank voor de samenwerking bij het opschrijven van hoofdstuk 7. Beste Maaike en Vera, dank jullie wel voor het meten van de stra lings-sterkte van alle fototherapielampen en daarna de verwerking van de gegevens voor het fototherapiestuk. Jullie hebben een prachtige ‘BART’ ontworpen die jullie daarna gelukkig ook zelf overal mee naartoe sleepten.

Graag wil ik alle andere co-auteurs danken voor de fijne samenwerking/I would like to thank all other co-authors: Claudio Tiribelli, Charles Ahlfors, Henkjan Verkade, Cas Weykamp en Christa Cobbaert.

Dear Prof. Ahlfors, thank you very much for sharing your knowledge on the complex matter of free bilirubin!

Beste Margreet & Andrea, we hebben heel wat meegemaakt in het lab met ons ‘bili-monster’. Fijn dat de metingen nu beter gaan. Het heeft bloed, zweet en tranen gekost. Dank voor de fijne samenwerking en veel succes in de toekomst! Beste Jan Dijks, dank je wel voor de technische ondersteuning.

Dankwoord 171

Dear Graham and Cassie, thank you very much for all the help with the measure-ment of free bilirubin levels.  

Alle neonatologen, fellows, arts-assistenten en (arts-)onderzoekers van de neo-natologie wil ik hartelijk danken voor de input, kritische blik en motiverende feedback die jullie geleverd hebben bij de researchbesprekingen. Ook heel hartelijk dank voor de ondersteuning en gezelligheid tijdens congressen.

Janette & Greetje, dank jullie wel voor alle hulp (en gezelligheid)! Jannie Tjassing en Aad van Mourik dank jullie wel voor de ondersteuning waar nodig.

Al het denkwerk heeft zich afgespeeld in de kelder van het Triadegebouw. Gelukkig zat ik daar niet alleen! Er zijn in de jaren een heleboel (arts-)onderzoekers gekomen en gegaan. Een paar wil ik bij naam noemen. Allereerst mijn kamergenootjes van kz0027: Karin, Tina, Bertine, Laura, Nicole, Annemieke en Djoeke. Wat hebben we veel meegemaakt met z’n allen! Dank jullie wel voor de gezelligheid maar ook voor de momenten van reflectie! Verder natuurlijk Elise Verhagen en Elise Roze dank jullie wel voor de gezelligheid tijdens congressen en de samenwerking tijdens de inclusieperiode en de follow up van onze studies. Ingrid, Eryn, Carianne, Hiltje, Paul, Menno, Martijn, Michelle, José en Chris Peter, dank jullie wel voor de gezelligheid. Janyte, dank je wel voor het prachtige ontwerp van de omslag!

Beste Koos, mede door jou heb ik ingezien dat ik mijn proefschrift echt af moest gaan maken. Dank je wel voor het leerzame eerste huisartsopleidingsjaar. Ik heb niet alleen veel over de huisartsgeneeskunde, maar ook veel over mezelf geleerd.

Lieve vrienden en familie, dank jullie wel voor jullie interesse in mijn proefschrift.

Lieve Janneke, dank je wel dat je mijn paranimf wilt zijn! Als pubers op de middelbare school, studenten in het Groningse studentenleven, Jip & Janneke in de boot en nu als moeders hebben we al een hele levensloop samen meegemaakt. Je bent me heel dierbaar. Dank je wel dat je er altijd voor me bent en altijd de juiste dingen weet te zeggen waardoor ik er weer tegenaan kan. Ik ben blij dat je in dit ‘huwelijk met de universiteit’ mijn bruidsmeisje wilt zijn!

Lieve Karin, je bent een waardevolle collega en vriendin geworden de afgelopen jaren. Ik ben ontzettend blij dat je eindelijk je welverdiende opleiding tot kinderarts bent begonnen. Dank je wel voor de gezellige tijd samen in de kelder en daarbuiten. Fijn dat je vandaag aan mijn zijde staat.

172

Lieve papa en mama, een paar regels in mijn boekje zijn natuurlijk nooit genoeg om mijn liefde en dank te uiten. Dit proefschrift was er nooit gekomen zonder jullie. Ik wil jullie in ieder geval heel erg danken voor jullie altijd veilige en warme thuishaven en jullie vertrouwen. Fijn dat jullie de laatste tijd zo vaak bij konden springen voor een warme maaltijd en hulp met de kinderen!

Lief broertje Joost, dank je wel voor je hulp bij het vinden van de juiste woorden.

Lieve Matthijs, je bent nooit gestopt te geloven dat het boekje eens af zou komen. De laatste loodjes waren zwaar maar ook waardevol. Ook dit hebben we samen gedaan! Dank je wel!

Lieve Noor en Nout, jullie zijn een verrijking van mijn leven. Heerlijk hoe alles van me af kon glijden zodra ik jullie smoeltjes zag. We gaan nog veel leuke dingen mee-maken samen! Lief klein kindje in mijn buik, wat ben ik benieuwd naar jou! Je hebt me vleugels gegeven tijdens de laatste loodjes van dit boekje!

1234567

9SD

8

173

Curriculum vitae

Deirdre Elisabeth van Imhoff werd op 16 januari 1981 geboren te Groningen. Tijdens haar middelbare schoolperiode heeft Deirdre een aantal jaren fanatiek wedstrijd geroeid. In 1999 deed ze haar eindexamen voorbereidend wetenschap-pelijk onderwijs aan het Maartenscollege te Groningen.

Aanvankelijk werd zij tweemaal uitgeloot voor de studie geneeskunde en haalde zij haar propedeuse psychologie aan de Rijksuniversiteit Groningen. In 2001 kon ze eindelijk beginnen aan de studie geneeskunde aan de Rijksuniversiteit van Groningen. Tijdens haar co-schappen startte zij met haar promotieonderzoek bij de onderafdeling neonatologie van het Beatrix Kinderziekenhuis van het UMC Groningen onder leiding van neonatologen prof. dr. A.F. Bos, dr. P.H. Dijk en dr. C.V. Hulzebos. Tijdens haar promotieonderzoek deed ze in 2008 artsexamen. Sinds maart 2011 is Deirdre in opleiding tot huisarts.

In 2009 trouwde ze met Matthijs Vader met wie ze in 2010 dochter Noor en in 2011 zoon Nout kreeg. In april 2013 verwachten ze hun 3e kind.

1234567

9SD

8

175

List of publications

Hulzebos CV, van Imhoff DE, Bos AF, Ahlfors CE, Verkade HJ, Dijk PH. Usefulness of the bilirubin/albumin ratio for predicting bilirubin-induced neurotoxicity in premature infants. Arch Dis Child Fetal Neonatal Ed 2008;93:F384–F388

van Imhoff DE, Dijk PH, Hulzebos CV; BARTrial-studiegroep van het Nederlands Neonatologisch Research Netwerk. Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a Dutch guideline. NTvG 2009;153:A94. Dutch.

van Imhoff DE, Dijk PH, Weykamp CW, Cobbaert CM, Hulzebos CV; On behalf of the BARTrial Study Group. Measurements of neonatal bilirubin and albumin concentrations: A need for improvement and quality control. Eur J Pediatr. 2011;170:977-82.

van Imhoff DE, Dijk PH, Hulzebos CV; BARTrial study group, Netherlands Neonatal Research Network. Uniform treatment thresholds for hyperbilirubinemia in preterm infants: Background and synopsis of a national guideline. Early Hum Dev 2011;87:521-5

van Imhoff DE, Hulzebos CV, van der Heide M, van den Belt VW, Vreman HJ, Dijk PH; the BARTrial Study Group. High variability and low irradiance of phototherapy devices in Dutch NICUs. Arch Dis Child Fetal Neonatal Ed. 2012 May 18. (Epub ahead of print)

Cuperus FJC, Schreuder AB, van Imhoff DE, Vitek L, Vanikova J, Konickova R, Ahlfors CE, Hulzebos CV, Verkade HJ. Beyond plasma bilirubin: the effects of phototherapy and albumin on brain bilirubin levels in Gunn rats. J Hepatol. 2012 Aug 21. (Epub ahead of print)

BARTrial Studygroup. A double-blind, randomized controlled trial on the bilirubin albumin ratio in jaundiced preterm infants Submitted

van Imhoff DE, Hulzebos CV, Bos AF, Dijk PH. Transcutaneous bilirubin measurements in preterm infants: Effects of phototherapy and treatment thresholds. Submitted

Othervan Imhoff DE, Cuperus FJC, Dijk PH, Tiribelli C, Hulzebos CV. Kernicterus,

Bilirubin Induced Neurological Dysfunction and New Treatments for Unconjugated Hyperbilirubinemia. G. Buonocore et al. (eds.), Neonatology. A Practical Approach to Neonatal Diseases. Springer-Verlag Italia 2011. Ch83:621-628