Journal of Perinatology (2015), 1–4 © 2015 Nature America, Inc. All rights reserved 0743-8346/15 www.nature.com/jp
ORIGINAL ARTICLE
Bilirubin production and hour-specific bilirubin levels VK Bhutani1, RJ Wong1, HJ Vreman1 and DK Stevenson1 on behalf of the Jaundice Multinational Study Group2 OBJECTIVE: We assessed the relative contributions of increased bilirubin production (indexed by end-tidal carbon monoxide (CO) concentrations, corrected for ambient CO (ETCOc)) to hour-specific total bilirubin (TB) levels in healthy late preterm and term newborns. STUDY DESIGN: Post hoc analyses of concurrent ETCOc and TB (at 30 ± 6 h of age) and follow-up TB levels at age 96 ± 12 h and up to 168 h after birth were performed in a cohort of 641 term and late preterm infants. RESULTS: Increased bilirubin production (hour-specific ETCOc ⩾ 1.7 p.p.m. at age 30 ± 6 h) was noted in ~ 80%, 42% and 32% of infants in the high-, intermediate- and low-risk TB zones, respectively. One infant with TB o40th percentile and ETCOc o 1.7 p.p.m. developed TB ⩾ 95th percentile at age 168 h, probably due to decreased bilirubin elimination. CONCLUSIONS: Infants in the high-risk quartile of the hour-specific bilirubin nomogram have a higher mean bilirubin production. Infants with TB levels ⩾ 95th percentile without increased bilirubin production have impaired bilirubin elimination. Journal of Perinatology advance online publication, 16 April 2015; doi:10.1038/jp.2015.32
INTRODUCTION In healthy term and late preterm newborns, peak serum/plasma total bilirubin (TB) levels often occur beyond 72 h of age and, more commonly, after discharge from the birthing hospital.1 In these apparently well infants, predischarge TB measurements can predict the likelihood of developing hyperbilirubinemia (TB ⩾ 95th percentile as defined on an hour-specific bilirubin nomogram2). As postnatal age increases, infants will generally remain in their predischarge-defined risk zones upon reaching a peak TB, which then declines gradually over time. Depending on the population, some infants may deviate from this normal pattern. This phenomenon is the result of a pathologic imbalance between bilirubin production and elimination. Increased production and/or decreased elimination of bilirubin have been considered as the contributing causes for neonatal jaundice and the subsequent development of hyperbilirubinemia. However, the relative contribution of these two processes to a healthy late preterm or term newborn’s predischarge hour-specific bilirubin percentile risk zone has not been previously characterized.3 Increased bilirubin production has been previously identified as a clinical risk factor for early-onset jaundice that is due to hemolysis secondary to Rh disease, ABO blood group incompatibility, cephalhematoma and bruising4,5 as well as in infants of diabetic mothers.6 Objective assessment of increased bilirubin production has been evidenced by increased levels of blood carboxyhemoglobin, end-tidal carbon monoxide (ETCOc), both corrected for ambient CO levels, or expired CO elimination rates.7 Earlier, Stevenson et al.6–9 had identified that while increased bilirubin production was a factor in some cases of early-onset hyperbilirubinemia, decreased bilirubin elimination could compound the bilirubin load.8 In addition, a normal gradual decline in bilirubin production was noted in infants without hemolysis as characterized by declining day-specific ETCOc values.10 Thus, for any infant, at any postnatal age during the first week after birth,
the development of hyperbilirubinemia depends on the balance of bilirubin production and its elimination.10 The purpose of this post hoc analysis of a previously reported study11 was to characterize the respective contributions of increased bilirubin production and, by inference, impaired bilirubin elimination to the infant’s assignment to a particular predischarge TB percentile risk zone. Thus, we measured concurrently the hourspecific TB and ETCOc levels in each infant to better understand the causal basis of neonatal hyperbilirubinemia and the subsequent risk of developing severe hyperbilirubinemia. METHODS Subject eligibility Infants were eligible for this post hoc analysis if they had participated in the study reported by Stevenson et al.11 In brief, infants were eligible to participate if their gestational age was ⩾ 35 weeks as determined by best obstetric estimate and if enrollment was accomplished within the first 36 h of age during 20 February 1998 to 22 February 1999. Infants were excluded who were jaundiced before enrollment, had severe congenital anomalies or any illness that required admission to the neonatal intensive care unit, or had pulmonary disease that required oxygen or ventilatory support. Notably, asymptomatic infants screened for sepsis based on risk criteria, but otherwise healthy, were included.
Hour-specific ETCOc nomogram Normal values are based on a healthy well-baby population (n = 2074) from the multinational study sites and Chicago (where the study was conducted by late Marguerite Herschel). Infants born to mothers with a self-declared history of smoking were excluded. All ETCOc values were measured using a CO-Stat™ End Tidal Breath Analyzer (Natus Medical, San Carlos, CA, USA now no longer commercially available) between 4- and 48-h postnatal age and then used to determine percentile distribution for age in hours. We selected the ETCOc value of ⩾ 1.7 p.p.m. as an index of excessive bilirubin production based on the paper by Stevenson et al.,11 which reported that ETCOc levels of hyperbilirubinemic infants were 1.8 ± 0.6 p.p.m. also using
1 Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA. Correspondence: Professor VK Bhutani, Department of Pediatrics, Stanford University School of Medicine, 750 Welch Road, Suite #315, Palo Alto, CA 94306, USA. E-mail: [email protected] 2 The members of Jaundice Multinational Study Group are listed before references. Received 26 November 2014; revised 13 February 2015; accepted 2 March 2015
Bilirubin production and hyperbilirubinemia VK Bhutani et al
2 Table 1.
Population demographics and characteristics
Characteristics
Mean
s.d. (range)
Birth weight (g) Age at first predischarge TB (h) Age at first ETCOc (h)
3308 28.3 28.3
483 (2070–5130) 3.1 (24.0–36.0) 3.4 (10.5–36.1)
n = 641
Percentage
309
48.2
242 85 25 221 68 139 37 92 35 22 227
37.8 13.3 3.9 34.5 10.6 21.7 5.8 14.4 5.5 3.4 35.4
Sample size
Figure 1. End-tidal carbon monoxide (CO) corrected for inhaled CO (ETCOc), percentiles and age-in-hours nomogram in 2074 healthy term and near term infants within 48 h of birth. (From Bhutani VK, Johnson LH, Stevenson DK, Herschel M, Group on behalf of the Jaundice Multinational Study Group. Pre-discharge (30 ± 6 h) diagnostic evaluation strategies for severe neonatal hyperbilirubinemia. Presented in abstract form at the American Academy of Pediatrics National Conference and Exhibition, 2001.
a CO-Stat device. This value corresponded to an ETCOc value at the 75th percentile in the ETCOc at 30 ± 6 h of age (Figure 1).
Study cohort with concurrent ETCO and TB measurements This cohort included all infants who met eligibility criteria (n = 793), for whom informed consent was obtained at their respective institution and whose ETCOc concentrations were tested using a CO-Stat device. Pre- and production samplers were used in this study, and Z-scores were used to normalize ETCOc values.11 Measurements of TB and ETCOc were performed at age 30 ± 6 h. Enrollment consisted of neonates from nine clinical sites (see Jaundice Multinational Study Group list of participants): four domestic and five international. Each study site enrolled eligible infants serially on a schedule determined by the circumstances imposed by clinical operations and personnel limitations at the respective institutions. Each site maintained an enrollment log identifying all infants eligible and selected for the study as well as the reasons for nonenrollment of any eligible neonate. Follow-up was also conducted at each site and the data were recorded in the research log. Exclusions were due to incomplete follow-up (n = 70), or use of a transcutaneous bilirubin only instead of TB (n = 82). Therefore, a total of 641 infants had both TB and ETCOc measurements at age 30 ± 6 h for this post hoc analysis. Pre- and postdischarge evaluation and outcome assessments were performed according to a protocol based on the hour-specific bilirubin nomogram, as detailed in the primary study.11
TB measurements Blood was collected and serum was separated and then analyzed for TB in each enrolled infant. Each site used its own clinical laboratory and method for all TB measurements. Infants who were not jaundiced had their first scheduled TB and ETCOc measurements performed at age 30 ± 6 h. All infants with TB ⩾ 95th percentile before discharge exited the study at this time. For the remainder, TB was performed at age 96 ± 12 h and subsequently at the clinician’s discretion10 until the infant attained age 168 h, at which point all infants exited the study to routine care of the clinician.
ETCOc measurements An ETCOc analyzer, with single-use disposable nasal sampler, was used to determine bilirubin production in all study infants.11,12 The device was calibrated locally every 30 days. The accuracy of the analyzer was ± 0.3 p.p. m. (or μl l− 1) or 10% of the reading (whichever was greater) for breathing rates between 10 and 60 min − 1 and 15% for breathing rates 460 min − 1. The reproducibility and accuracy of ETCOc measurements using this device has been previously described in detail.11,13 The device used side-stream sampling to draw nasal air continuously through a sampler at a rate of 60 ml min −1. The sampler was made of a clear polymer with an inner and outer diameter of 0.8 and 1.5 mm, respectively. Adhesive wings allowed for a maximal nasal insertion of 6 mm. Because of the design of this device, breath control was not essential to maintain the integrity of the end-tidal Journal of Perinatology (2015), 1 – 4
Male Race Asian Black Hispanic White Others Cesarean section Rh-negative mothers ABO incompatibility Bruising Cephalhematoma Exclusive breastfeeding
Abbreviations: ETCOc, end-tidal carbon monoxide (CO) concentrations, corrected for ambient CO; TB, total bilirubin. Data are presented for study patients with both predischarge TB and ETCOc measurements who complete follow-up during the first week. Demographic data was incomplete in two infants.
breath sample of gas; further corroborated by the concurrent measure of end-tidal carbon dioxide concentrations.
Statistics Data collected at age 30 ± 6 h (predischarge) were stratified by the risk status on the hour-specific bilirubin nomogram. Thus, infants were grouped as being ⩾ 95th percentile (high risk), 40th to 94th percentile (intermediate risk) and o40th percentile (low risk) for age in hours. The relationship between increasing risk zones of hour-specific TB and ETCOc was determined by linear regression. Furthermore, differences in ETCOc for each risk zone were compared by Student’s t-test. Values of ETCOc ⩾ 1.7 p.p.m. at age 30 ± 6 h were above the 75th percentile for age in hours and regarded as indicative of increased bilirubin production.
RESULTS Normal values are based on ETCOc values measured in a healthy well-baby population (n = 2074) between 4- and 48-h postnatal age. Percentile distribution for age in hours is shown graphically in Figure 1. The ETCOc values at the 75th percentile level on the ETCOc nomogram, as an index for increased bilirubin production are illustrated. Birth weight, gestational age, racial and perinatal characteristics of the study cohort with concurrent TB and ETCO values, evaluated before discharge and at follow-up (n = 641), are listed in Table 1. The distribution of predischarge study patients according to their bilirubin risk status and the concurrent ETCOc values (mean ± s.d., range) at age 30 ± 6 h is shown according to the TB percentile tracks in Table 2. Approximately 80% (12/15) of infants in the high-risk group had an elevated ETCOc with a mean of 2.3 ± 0.8 vs 1.4 ± 0.5 p.p.m. in the low-risk group (Po 0.01). Only one infant with an TB o 40th percentile and ETCOc o 1.7 p.p.m. subsequently developed an TB ⩾ 95th percentile (Figure 2). DISCUSSION This post hoc analysis of previously reported data11 provides a basis for further understanding of how increased production and decreased elimination of bilirubin contribute to the development of neonatal hyperbilirubinemia. In this analysis, we found that © 2015 Nature America, Inc.
Bilirubin production and hyperbilirubinemia VK Bhutani et al
3 Table 2. Predischarge hour-specific serum/plasma total bilirubin (TB) and end-tidal carbon monoxide (CO), corrected for inhaled CO (ETCOc) values of study population at each risk zone Risk zone High Risk High intermediate Low intermediate Low risk Total
Hour-specific TB percentiles ⩾ 95th 76–94th 40–75th o40th
Number (percentage) of infants in each risk zone 15 117 238 271 641
(2.3) (18.3) (37.1) (42.3) (100)
Mean ETCOc ± s.d. (range) 2.3 ± 0.8* 1.6 ± 0.5 1.5 ± 0.5 1.4 ± 0.5 1.5 ±0.5
(1.1–2.9) (0.4–2.7) (0.2–3.7) (0.2–6.3) (0.2–6.3)
Percentage (number) of infants with ETCOc ⩾ 1.7 p.p.m. 80.0 53.9 36.6 32.0 38.8
(12/15) (63/117) (87/238) (86/269) (248/639)
Abbreviations: ETCOc, end-tidal carbon monoxide (CO) concentrations, corrected for ambient CO; TB, total bilirubin. Data are for study infants with both predischarge TB and ETCOc measurements (*Po0.01).
Figure 2. Case report of a Chinese male infant cared for at the Pamela Youde Nethersole Eastern Hospital (Hong Kong, China) with a slowly progressive severe nonhemolytic hyperbilirubinemia. This was a term male infant with a birth weight (BW) of 3886 g who was delivered by a spontaneous vaginal delivery. He was breastfed and supplemented with bottle feeds. There was no ABO or Rh incompatibility, Coombs’ test was negative, and there were no bruises or cephalhematoma. ETCOc level measured at age 23 h was 0.4 p.p.m. and not indicative of increased bilirubin production. Glucose-6-phosphate dehydrogenase (G6PD) activity level was unknown. The infant was lost to follow-up after 1 month of age. However, this pattern depicts infrequent neonatal hyperbilirubinemia that is strongly suggestive of an inadequate or delayed bilirubin elimination disorder.
increased bilirubin production (that is, ETCOc ⩾ 1.7 p.p.m.) was observed in 80% of infants with TB ⩾ 95th percentile, 42% of those with TB between 40th and 95th percentiles, and 32% of those with TB o40th percentile. Thus, increased bilirubin production is an important contributing factor for TB levels being ⩾ 95th percentile. However, we also observed that impaired bilirubin elimination was the predominant contributing factor in infants with TB o 95th percentile, many of whom are low-bilirubin producers, and, in most cases, reflect the normal physiologic circumstance of the transitional period after birth. Moreover, better bilirubin elimination may also account for the lack of severe hyperbilirubinemia in some high bilirubin producers, reducing the positive predictive accuracy of ETCOc measurements alone. Neonatal hyperbilirubinemia, not attributable to increased bilirubin production, is likely to occur in infants with increased enterohepatic circulation (as with starvation, failed initiation of breastfeeding, or decreased gastrointestinal motility), and in infants with decreased uridine diphosphate glucuronosyltransferase enzyme activity, such as those associated with UG1TA1 gene polymorphisms14–16 as suggested by the relationship in Figure 2. Recent evidence have also shown that microsatellite polymorphisms in the HO-1 promoter region called (GT)n repeats are associated with varying levels of HO-1 expression, with infants © 2015 Nature America, Inc.
having shorter (GT)n repeat lengths being high HO-1 expressors.17 Therefore, these infants may have high bilirubin production rates, and together with an impaired bilirubin conjugation capacity, will be at high risk for developing severe neonatal hyperbilirubinemia in the context of hemolysis. This study affirms the previous report2 that infants with TB o40th percentile are at minimal risk for developing subsequent severe hyperbilirubinemia, even when they may have increased bilirubin production because most infants have rapidly improving conjugation after birth. In general, infants with ‘normal’ bilirubin production (that is, ETCOc o 1.7 p.p.m.) and TB o 40th percentile may be characterized as having ‘benign’ hyperbilirubinemia with the least likelihood of developing severe hyperbilirubinemia. In the multinational study by Stevenson et al.,11 ETCOc levels at 30 ± 6 h for the total population was 1.48 ± 0.49 p.p.m., whereas, those for nonhyperbilirubinemic and hyperbilirubinemic infants were 1.45 ± 0.47 and 1.81 ± 0.59 p.p.m., respectively. Seventy-six percent (92 of 120) of hyperbilirubinemic infants had ETCOc greater than the population mean. One infant (described in Figure 2) who had normal bilirubin production (TB o 40th percentile at age 30 ± 6 h), and ultimately developed a TB ⩾ 95th percentile before age 168 h, was easily identified as having a problem with bilirubin elimination and was evaluated further for a possible conjugation defect. Such infants are unusual, but can be observed to be deviating from initial lowrisk zones (percentile tracks) in the presence of normal bilirubin production, thus suggesting inadequate bilirubin elimination. This is important information when planning follow-up for infants that have TB levels within the normal range, but TB levels are moving into higher percentile risk zones with normal bilirubin production. Predischarge diagnostic evaluation with both ETCOc and TB and follow-up of such infants would allow for further targeted and genetic investigations of such infants for bilirubin conjugation disorders. Maisels and Kring18 had reported that there are sustained or increased day-specific ETCOc levels in jaundiced infants as compared with decreasing ETCOc levels in control infants during the first 4 days after birth. Their data suggest that increased heme catabolism is an important mechanism responsible for the development of this hyperbilirubinemia. Predischarge assessment of risk for hyperbilirubinemia according to the 2004 American Academy of Pediatrics practice guideline1 should continue to decrease the risk of developing severe hyperbilirubinemia and, more importantly, kernicterus. Clinical, epidemiological and demographic risk factors that predispose infants to such an adverse outcome have been identified.1,19 Individually, these risk factors may be of practical use for the practicing clinician because they have some clinical predictive value and rule-based predictive risk scores that combine both clinical and laboratory indices, similar to those described by Newman et al.20 and Keren et al.21 In addition to measurements of both hour-specific TB and ETCOc to identify infants with bilirubin Journal of Perinatology (2015), 1 – 4
Bilirubin production and hyperbilirubinemia VK Bhutani et al
4
production and risk of hemolysis,22 a clinician may deduce those with delayed bilirubin elimination. Infants with imbalances in conjugation defects need close follow-up and appropriate, and possibly more targeted, therapies (when such options exist) can be performed. Importantly, the hyperbilirubinemia in a substantial number of infants reported in the Pilot Kernicterus Registry could only be attributed to unknown causes.23,24 Knowledge about bilirubin production in association with an hour-specific TB measurement might allow for the diagnostic characterization of such an ‘idiopathic’ group by identifying infants with increased production or decreased elimination of bilirubin as the major contributing cause to their late onset neonatal hyperbilirubinemia. CONFLICT OF INTEREST The authors declare no conflict of interest.
JAUNDICE MULTINATIONAL STUDY GROUP David K Stevenson, Ronald J Wong, Hendrik J Vreman and James R MacMahon (Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA); Avroy A Fanaroff (Department of Neonatology, Rainbow Babies’ and Children’s Hospital, Cleveland, OH, USA); M Jeffrey Maisels (Department of Pediatrics, Wm. Beaumont Hospital, Royal Oak, MI, USA); Betty WY Young (Department of Pediatrics, Pamela Youde Nethersole Eastern Hospital, Hong Kong, China); Chap Y Yeung (Department of Pediatrics, Queen Mary Hospital/Tsan Yuk Maternity Hospital, Hong Kong, China); Daniel S Seidman and Rena Gale (Department of Pediatrics, Bikur Cholim Hospital, Jerusalem, Israel); William Oh (Department of Pediatrics, Women’s and Infants Hospital, Providence, RI, USA); Vinod K Bhutani and Lois H Johnson (Section on Newborn Pediatrics, Pennsylvania Hospital, Philadelphia, PA, USA); Michael Kaplan and Cathy Hammerman (Department of Neonatology, Shaare Zedek Medical Center, Jerusalem, Israel); Hajime Nakamura (Department of Pediatrics, Kobe University Graduate School of Medicine, Kobe, Japan). REFERENCES 1 American Academy of Pediatrics. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004; 114: 297–316. 2 Bhutani VK, Johnson L, Sivieri EM. Predictive ability of a predischarge hour-specific serum bilirubin for subsequent significant hyperbilirubinemia in healthy term and near-term newborns. Pediatrics 1999; 103: 6–14. 3 Kaplan M, Muraca M, Hammerman C, Rubaltelli FF, Vilei MT, Vreman HJ et al. Imbalance between production and conjugation of bilirubin: A fundamental concept in the mechanism of neonatal jaundice. Pediatrics 2002; 110: e47–e51. 4 Maisels MJ, Pathak A, Nelson NM, Nathan DG, Smith CA. Endogenous production of carbon monoxide in normal and erythroblastotic newborn infants. J Clin Invest 1971; 50: 1–8.
Journal of Perinatology (2015), 1 – 4
5 Stevenson DK, Vreman HJ. Carbon monoxide and bilirubin production in neonates. Pediatrics 1997; 100: 252–254. 6 Stevenson DK, Bartoletti AL, Ostrander CR, Johnson JD. Pulmonary excretion of carbon monoxide in the human infant as an index of bilirubin production. II. Infants of diabetic mothers. J Pediatr 1979; 94: 956–958. 7 Stevenson DK, Bartoletti AL, Ostrander CR, Johnson JD. Pulmonary excretion of carbon monoxide in the human newborn infant as an index of bilirubin production: III. Measurement of pulmonary excretion of carbon monoxide after the first postnatal week in premature infants. Pediatrics 1979; 64: 598–600. 8 Stevenson DK, Ostrander CR, Hopper AO, Cohen RS, Johnson JD. Pulmonary excretion of carbon monoxide as an index of bilirubin production. IIa. Evidence for possible delayed clearance of bilirubin in infants of diabetic mothers. J Pediatr 1981; 98: 822–824. 9 Stevenson DK, Vreman HJ, Wong RJ, Contag CH. Carbon monoxide and bilirubin production in neonates. Semin Perinatol 2001; 25: 85–93. 10 Balaraman V, Pelke S, DiMauro S, Cheung S, Stevenson DK, Easa D. End-tidal carbon monoxide in newborn infants: observations during the 1st week of life. Biol Neonate 1995; 67: 182–185. 11 Stevenson DK, Fanaroff AA, Maisels MJ, Young BW, Wong RJ, Vreman HJ et al. Prediction of hyperbilirubinemia in near-term and term infants. Pediatrics 2001; 108: 31–39. 12 Bhutani VK, Stark AR, Lazzeroni LC, Poland R, Gourley GR, Kazmierczak S et al. Predischarge screening for severe neonatal hyperbilirubinemia identifies infants who need phototherapy. J Pediatr 2013; 162: 477–482 e471. 13 Vreman HJ, Wong RJ, Harmatz P, Fanaroff AA, Berman B, Stevenson DK. Validation of the natus CO-Stat end tidal breath analyzer in children and adults. J Clin Monit Comput 1999; 15: 421–427. 14 Beutler E, Gelbart T, Demina A. Racial variability in the UDPglucuronosyltransferase 1 (UGT1A1) promoter: a balanced polymorphism for regulation of bilirubin metabolism? Proc Natl Acad Sci USA 1998; 95: 8170–8174. 15 Bosma PJ, Chowdhury JR, Bakker C, Gantla S, de Boer A, Oostra BA et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyltransferase 1 in Gilbert's syndrome. N Engl J Med 1995; 333: 1171–1175. 16 Kaplan M, Renbaum P, Levy-Lahad E, Hammerman C, Lahad A, Beutler E. Gilbert syndrome and glucose-6-phosphate dehydrogenase deficiency: a dosedependent genetic interaction crucial to neonatal hyperbilirubinemia. Proc Natl Acad Sci USA 1997; 94: 12128–12132. 17 Shibahara S, Kitamuro T, Takahashi K. Heme degradation and human disease: diversity is the soul of life. Antioxid Redox Signal 2002; 4: 593–602. 18 Maisels MJ, Kring E. The contribution of hemolysis to early jaundice in normal newborns. Pediatrics 2006; 118: 276–279. 19 Dennery PA, Seidman DS, Stevenson DK. Neonatal hyperbilirubinemia. N Engl J Med 2001; 344: 581–590. 20 Newman TB, Xiong B, Gonzales VM, Escobar GJ. Prediction and prevention of extreme neonatal hyperbilirubinemia in a mature health maintenance organization. Arch Pediatr Adolesc Med 2000; 154: 1140–1147. 21 Keren R, Bhutani VK, Luan X, Nihtianova S, Cnaan A, Schwartz JS. Identifying newborns at risk of significant hyperbilirubinaemia: a comparison of two recommended approaches. Arch Dis Child 2005; 90: 415–421. 22 Tidmarsh GF, Wong RJ, Stevenson DK. End-tidal carbon monoxide and hemolysis. J Perinatol 2014; 34: 577–581. 23 Bhutani VK, Johnson LH, Jeffrey Maisels M, Newman TB, Phibbs C, Stark AR et al. Kernicterus: epidemiological strategies for its prevention through systems-based approaches. J Perinatol 2004; 24: 650–662. 24 Johnson LH, Bhutani VK, Brown AK. System-based approach to management of neonatal jaundice and prevention of kernicterus. J Pediatr 2002; 140: 396–403.
© 2015 Nature America, Inc.
ORIGINAL ARTICLE
Bilirubin production and hour-specific bilirubin levels VK Bhutani1, RJ Wong1, HJ Vreman1 and DK Stevenson1 on behalf of the Jaundice Multinational Study Group2 OBJECTIVE: We assessed the relative contributions of increased bilirubin production (indexed by end-tidal carbon monoxide (CO) concentrations, corrected for ambient CO (ETCOc)) to hour-specific total bilirubin (TB) levels in healthy late preterm and term newborns. STUDY DESIGN: Post hoc analyses of concurrent ETCOc and TB (at 30 ± 6 h of age) and follow-up TB levels at age 96 ± 12 h and up to 168 h after birth were performed in a cohort of 641 term and late preterm infants. RESULTS: Increased bilirubin production (hour-specific ETCOc ⩾ 1.7 p.p.m. at age 30 ± 6 h) was noted in ~ 80%, 42% and 32% of infants in the high-, intermediate- and low-risk TB zones, respectively. One infant with TB o40th percentile and ETCOc o 1.7 p.p.m. developed TB ⩾ 95th percentile at age 168 h, probably due to decreased bilirubin elimination. CONCLUSIONS: Infants in the high-risk quartile of the hour-specific bilirubin nomogram have a higher mean bilirubin production. Infants with TB levels ⩾ 95th percentile without increased bilirubin production have impaired bilirubin elimination. Journal of Perinatology advance online publication, 16 April 2015; doi:10.1038/jp.2015.32
INTRODUCTION In healthy term and late preterm newborns, peak serum/plasma total bilirubin (TB) levels often occur beyond 72 h of age and, more commonly, after discharge from the birthing hospital.1 In these apparently well infants, predischarge TB measurements can predict the likelihood of developing hyperbilirubinemia (TB ⩾ 95th percentile as defined on an hour-specific bilirubin nomogram2). As postnatal age increases, infants will generally remain in their predischarge-defined risk zones upon reaching a peak TB, which then declines gradually over time. Depending on the population, some infants may deviate from this normal pattern. This phenomenon is the result of a pathologic imbalance between bilirubin production and elimination. Increased production and/or decreased elimination of bilirubin have been considered as the contributing causes for neonatal jaundice and the subsequent development of hyperbilirubinemia. However, the relative contribution of these two processes to a healthy late preterm or term newborn’s predischarge hour-specific bilirubin percentile risk zone has not been previously characterized.3 Increased bilirubin production has been previously identified as a clinical risk factor for early-onset jaundice that is due to hemolysis secondary to Rh disease, ABO blood group incompatibility, cephalhematoma and bruising4,5 as well as in infants of diabetic mothers.6 Objective assessment of increased bilirubin production has been evidenced by increased levels of blood carboxyhemoglobin, end-tidal carbon monoxide (ETCOc), both corrected for ambient CO levels, or expired CO elimination rates.7 Earlier, Stevenson et al.6–9 had identified that while increased bilirubin production was a factor in some cases of early-onset hyperbilirubinemia, decreased bilirubin elimination could compound the bilirubin load.8 In addition, a normal gradual decline in bilirubin production was noted in infants without hemolysis as characterized by declining day-specific ETCOc values.10 Thus, for any infant, at any postnatal age during the first week after birth,
the development of hyperbilirubinemia depends on the balance of bilirubin production and its elimination.10 The purpose of this post hoc analysis of a previously reported study11 was to characterize the respective contributions of increased bilirubin production and, by inference, impaired bilirubin elimination to the infant’s assignment to a particular predischarge TB percentile risk zone. Thus, we measured concurrently the hourspecific TB and ETCOc levels in each infant to better understand the causal basis of neonatal hyperbilirubinemia and the subsequent risk of developing severe hyperbilirubinemia. METHODS Subject eligibility Infants were eligible for this post hoc analysis if they had participated in the study reported by Stevenson et al.11 In brief, infants were eligible to participate if their gestational age was ⩾ 35 weeks as determined by best obstetric estimate and if enrollment was accomplished within the first 36 h of age during 20 February 1998 to 22 February 1999. Infants were excluded who were jaundiced before enrollment, had severe congenital anomalies or any illness that required admission to the neonatal intensive care unit, or had pulmonary disease that required oxygen or ventilatory support. Notably, asymptomatic infants screened for sepsis based on risk criteria, but otherwise healthy, were included.
Hour-specific ETCOc nomogram Normal values are based on a healthy well-baby population (n = 2074) from the multinational study sites and Chicago (where the study was conducted by late Marguerite Herschel). Infants born to mothers with a self-declared history of smoking were excluded. All ETCOc values were measured using a CO-Stat™ End Tidal Breath Analyzer (Natus Medical, San Carlos, CA, USA now no longer commercially available) between 4- and 48-h postnatal age and then used to determine percentile distribution for age in hours. We selected the ETCOc value of ⩾ 1.7 p.p.m. as an index of excessive bilirubin production based on the paper by Stevenson et al.,11 which reported that ETCOc levels of hyperbilirubinemic infants were 1.8 ± 0.6 p.p.m. also using
1 Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA. Correspondence: Professor VK Bhutani, Department of Pediatrics, Stanford University School of Medicine, 750 Welch Road, Suite #315, Palo Alto, CA 94306, USA. E-mail: [email protected] 2 The members of Jaundice Multinational Study Group are listed before references. Received 26 November 2014; revised 13 February 2015; accepted 2 March 2015
Bilirubin production and hyperbilirubinemia VK Bhutani et al
2 Table 1.
Population demographics and characteristics
Characteristics
Mean
s.d. (range)
Birth weight (g) Age at first predischarge TB (h) Age at first ETCOc (h)
3308 28.3 28.3
483 (2070–5130) 3.1 (24.0–36.0) 3.4 (10.5–36.1)
n = 641
Percentage
309
48.2
242 85 25 221 68 139 37 92 35 22 227
37.8 13.3 3.9 34.5 10.6 21.7 5.8 14.4 5.5 3.4 35.4
Sample size
Figure 1. End-tidal carbon monoxide (CO) corrected for inhaled CO (ETCOc), percentiles and age-in-hours nomogram in 2074 healthy term and near term infants within 48 h of birth. (From Bhutani VK, Johnson LH, Stevenson DK, Herschel M, Group on behalf of the Jaundice Multinational Study Group. Pre-discharge (30 ± 6 h) diagnostic evaluation strategies for severe neonatal hyperbilirubinemia. Presented in abstract form at the American Academy of Pediatrics National Conference and Exhibition, 2001.
a CO-Stat device. This value corresponded to an ETCOc value at the 75th percentile in the ETCOc at 30 ± 6 h of age (Figure 1).
Study cohort with concurrent ETCO and TB measurements This cohort included all infants who met eligibility criteria (n = 793), for whom informed consent was obtained at their respective institution and whose ETCOc concentrations were tested using a CO-Stat device. Pre- and production samplers were used in this study, and Z-scores were used to normalize ETCOc values.11 Measurements of TB and ETCOc were performed at age 30 ± 6 h. Enrollment consisted of neonates from nine clinical sites (see Jaundice Multinational Study Group list of participants): four domestic and five international. Each study site enrolled eligible infants serially on a schedule determined by the circumstances imposed by clinical operations and personnel limitations at the respective institutions. Each site maintained an enrollment log identifying all infants eligible and selected for the study as well as the reasons for nonenrollment of any eligible neonate. Follow-up was also conducted at each site and the data were recorded in the research log. Exclusions were due to incomplete follow-up (n = 70), or use of a transcutaneous bilirubin only instead of TB (n = 82). Therefore, a total of 641 infants had both TB and ETCOc measurements at age 30 ± 6 h for this post hoc analysis. Pre- and postdischarge evaluation and outcome assessments were performed according to a protocol based on the hour-specific bilirubin nomogram, as detailed in the primary study.11
TB measurements Blood was collected and serum was separated and then analyzed for TB in each enrolled infant. Each site used its own clinical laboratory and method for all TB measurements. Infants who were not jaundiced had their first scheduled TB and ETCOc measurements performed at age 30 ± 6 h. All infants with TB ⩾ 95th percentile before discharge exited the study at this time. For the remainder, TB was performed at age 96 ± 12 h and subsequently at the clinician’s discretion10 until the infant attained age 168 h, at which point all infants exited the study to routine care of the clinician.
ETCOc measurements An ETCOc analyzer, with single-use disposable nasal sampler, was used to determine bilirubin production in all study infants.11,12 The device was calibrated locally every 30 days. The accuracy of the analyzer was ± 0.3 p.p. m. (or μl l− 1) or 10% of the reading (whichever was greater) for breathing rates between 10 and 60 min − 1 and 15% for breathing rates 460 min − 1. The reproducibility and accuracy of ETCOc measurements using this device has been previously described in detail.11,13 The device used side-stream sampling to draw nasal air continuously through a sampler at a rate of 60 ml min −1. The sampler was made of a clear polymer with an inner and outer diameter of 0.8 and 1.5 mm, respectively. Adhesive wings allowed for a maximal nasal insertion of 6 mm. Because of the design of this device, breath control was not essential to maintain the integrity of the end-tidal Journal of Perinatology (2015), 1 – 4
Male Race Asian Black Hispanic White Others Cesarean section Rh-negative mothers ABO incompatibility Bruising Cephalhematoma Exclusive breastfeeding
Abbreviations: ETCOc, end-tidal carbon monoxide (CO) concentrations, corrected for ambient CO; TB, total bilirubin. Data are presented for study patients with both predischarge TB and ETCOc measurements who complete follow-up during the first week. Demographic data was incomplete in two infants.
breath sample of gas; further corroborated by the concurrent measure of end-tidal carbon dioxide concentrations.
Statistics Data collected at age 30 ± 6 h (predischarge) were stratified by the risk status on the hour-specific bilirubin nomogram. Thus, infants were grouped as being ⩾ 95th percentile (high risk), 40th to 94th percentile (intermediate risk) and o40th percentile (low risk) for age in hours. The relationship between increasing risk zones of hour-specific TB and ETCOc was determined by linear regression. Furthermore, differences in ETCOc for each risk zone were compared by Student’s t-test. Values of ETCOc ⩾ 1.7 p.p.m. at age 30 ± 6 h were above the 75th percentile for age in hours and regarded as indicative of increased bilirubin production.
RESULTS Normal values are based on ETCOc values measured in a healthy well-baby population (n = 2074) between 4- and 48-h postnatal age. Percentile distribution for age in hours is shown graphically in Figure 1. The ETCOc values at the 75th percentile level on the ETCOc nomogram, as an index for increased bilirubin production are illustrated. Birth weight, gestational age, racial and perinatal characteristics of the study cohort with concurrent TB and ETCO values, evaluated before discharge and at follow-up (n = 641), are listed in Table 1. The distribution of predischarge study patients according to their bilirubin risk status and the concurrent ETCOc values (mean ± s.d., range) at age 30 ± 6 h is shown according to the TB percentile tracks in Table 2. Approximately 80% (12/15) of infants in the high-risk group had an elevated ETCOc with a mean of 2.3 ± 0.8 vs 1.4 ± 0.5 p.p.m. in the low-risk group (Po 0.01). Only one infant with an TB o 40th percentile and ETCOc o 1.7 p.p.m. subsequently developed an TB ⩾ 95th percentile (Figure 2). DISCUSSION This post hoc analysis of previously reported data11 provides a basis for further understanding of how increased production and decreased elimination of bilirubin contribute to the development of neonatal hyperbilirubinemia. In this analysis, we found that © 2015 Nature America, Inc.
Bilirubin production and hyperbilirubinemia VK Bhutani et al
3 Table 2. Predischarge hour-specific serum/plasma total bilirubin (TB) and end-tidal carbon monoxide (CO), corrected for inhaled CO (ETCOc) values of study population at each risk zone Risk zone High Risk High intermediate Low intermediate Low risk Total
Hour-specific TB percentiles ⩾ 95th 76–94th 40–75th o40th
Number (percentage) of infants in each risk zone 15 117 238 271 641
(2.3) (18.3) (37.1) (42.3) (100)
Mean ETCOc ± s.d. (range) 2.3 ± 0.8* 1.6 ± 0.5 1.5 ± 0.5 1.4 ± 0.5 1.5 ±0.5
(1.1–2.9) (0.4–2.7) (0.2–3.7) (0.2–6.3) (0.2–6.3)
Percentage (number) of infants with ETCOc ⩾ 1.7 p.p.m. 80.0 53.9 36.6 32.0 38.8
(12/15) (63/117) (87/238) (86/269) (248/639)
Abbreviations: ETCOc, end-tidal carbon monoxide (CO) concentrations, corrected for ambient CO; TB, total bilirubin. Data are for study infants with both predischarge TB and ETCOc measurements (*Po0.01).
Figure 2. Case report of a Chinese male infant cared for at the Pamela Youde Nethersole Eastern Hospital (Hong Kong, China) with a slowly progressive severe nonhemolytic hyperbilirubinemia. This was a term male infant with a birth weight (BW) of 3886 g who was delivered by a spontaneous vaginal delivery. He was breastfed and supplemented with bottle feeds. There was no ABO or Rh incompatibility, Coombs’ test was negative, and there were no bruises or cephalhematoma. ETCOc level measured at age 23 h was 0.4 p.p.m. and not indicative of increased bilirubin production. Glucose-6-phosphate dehydrogenase (G6PD) activity level was unknown. The infant was lost to follow-up after 1 month of age. However, this pattern depicts infrequent neonatal hyperbilirubinemia that is strongly suggestive of an inadequate or delayed bilirubin elimination disorder.
increased bilirubin production (that is, ETCOc ⩾ 1.7 p.p.m.) was observed in 80% of infants with TB ⩾ 95th percentile, 42% of those with TB between 40th and 95th percentiles, and 32% of those with TB o40th percentile. Thus, increased bilirubin production is an important contributing factor for TB levels being ⩾ 95th percentile. However, we also observed that impaired bilirubin elimination was the predominant contributing factor in infants with TB o 95th percentile, many of whom are low-bilirubin producers, and, in most cases, reflect the normal physiologic circumstance of the transitional period after birth. Moreover, better bilirubin elimination may also account for the lack of severe hyperbilirubinemia in some high bilirubin producers, reducing the positive predictive accuracy of ETCOc measurements alone. Neonatal hyperbilirubinemia, not attributable to increased bilirubin production, is likely to occur in infants with increased enterohepatic circulation (as with starvation, failed initiation of breastfeeding, or decreased gastrointestinal motility), and in infants with decreased uridine diphosphate glucuronosyltransferase enzyme activity, such as those associated with UG1TA1 gene polymorphisms14–16 as suggested by the relationship in Figure 2. Recent evidence have also shown that microsatellite polymorphisms in the HO-1 promoter region called (GT)n repeats are associated with varying levels of HO-1 expression, with infants © 2015 Nature America, Inc.
having shorter (GT)n repeat lengths being high HO-1 expressors.17 Therefore, these infants may have high bilirubin production rates, and together with an impaired bilirubin conjugation capacity, will be at high risk for developing severe neonatal hyperbilirubinemia in the context of hemolysis. This study affirms the previous report2 that infants with TB o40th percentile are at minimal risk for developing subsequent severe hyperbilirubinemia, even when they may have increased bilirubin production because most infants have rapidly improving conjugation after birth. In general, infants with ‘normal’ bilirubin production (that is, ETCOc o 1.7 p.p.m.) and TB o 40th percentile may be characterized as having ‘benign’ hyperbilirubinemia with the least likelihood of developing severe hyperbilirubinemia. In the multinational study by Stevenson et al.,11 ETCOc levels at 30 ± 6 h for the total population was 1.48 ± 0.49 p.p.m., whereas, those for nonhyperbilirubinemic and hyperbilirubinemic infants were 1.45 ± 0.47 and 1.81 ± 0.59 p.p.m., respectively. Seventy-six percent (92 of 120) of hyperbilirubinemic infants had ETCOc greater than the population mean. One infant (described in Figure 2) who had normal bilirubin production (TB o 40th percentile at age 30 ± 6 h), and ultimately developed a TB ⩾ 95th percentile before age 168 h, was easily identified as having a problem with bilirubin elimination and was evaluated further for a possible conjugation defect. Such infants are unusual, but can be observed to be deviating from initial lowrisk zones (percentile tracks) in the presence of normal bilirubin production, thus suggesting inadequate bilirubin elimination. This is important information when planning follow-up for infants that have TB levels within the normal range, but TB levels are moving into higher percentile risk zones with normal bilirubin production. Predischarge diagnostic evaluation with both ETCOc and TB and follow-up of such infants would allow for further targeted and genetic investigations of such infants for bilirubin conjugation disorders. Maisels and Kring18 had reported that there are sustained or increased day-specific ETCOc levels in jaundiced infants as compared with decreasing ETCOc levels in control infants during the first 4 days after birth. Their data suggest that increased heme catabolism is an important mechanism responsible for the development of this hyperbilirubinemia. Predischarge assessment of risk for hyperbilirubinemia according to the 2004 American Academy of Pediatrics practice guideline1 should continue to decrease the risk of developing severe hyperbilirubinemia and, more importantly, kernicterus. Clinical, epidemiological and demographic risk factors that predispose infants to such an adverse outcome have been identified.1,19 Individually, these risk factors may be of practical use for the practicing clinician because they have some clinical predictive value and rule-based predictive risk scores that combine both clinical and laboratory indices, similar to those described by Newman et al.20 and Keren et al.21 In addition to measurements of both hour-specific TB and ETCOc to identify infants with bilirubin Journal of Perinatology (2015), 1 – 4
Bilirubin production and hyperbilirubinemia VK Bhutani et al
4
production and risk of hemolysis,22 a clinician may deduce those with delayed bilirubin elimination. Infants with imbalances in conjugation defects need close follow-up and appropriate, and possibly more targeted, therapies (when such options exist) can be performed. Importantly, the hyperbilirubinemia in a substantial number of infants reported in the Pilot Kernicterus Registry could only be attributed to unknown causes.23,24 Knowledge about bilirubin production in association with an hour-specific TB measurement might allow for the diagnostic characterization of such an ‘idiopathic’ group by identifying infants with increased production or decreased elimination of bilirubin as the major contributing cause to their late onset neonatal hyperbilirubinemia. CONFLICT OF INTEREST The authors declare no conflict of interest.
JAUNDICE MULTINATIONAL STUDY GROUP David K Stevenson, Ronald J Wong, Hendrik J Vreman and James R MacMahon (Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA); Avroy A Fanaroff (Department of Neonatology, Rainbow Babies’ and Children’s Hospital, Cleveland, OH, USA); M Jeffrey Maisels (Department of Pediatrics, Wm. Beaumont Hospital, Royal Oak, MI, USA); Betty WY Young (Department of Pediatrics, Pamela Youde Nethersole Eastern Hospital, Hong Kong, China); Chap Y Yeung (Department of Pediatrics, Queen Mary Hospital/Tsan Yuk Maternity Hospital, Hong Kong, China); Daniel S Seidman and Rena Gale (Department of Pediatrics, Bikur Cholim Hospital, Jerusalem, Israel); William Oh (Department of Pediatrics, Women’s and Infants Hospital, Providence, RI, USA); Vinod K Bhutani and Lois H Johnson (Section on Newborn Pediatrics, Pennsylvania Hospital, Philadelphia, PA, USA); Michael Kaplan and Cathy Hammerman (Department of Neonatology, Shaare Zedek Medical Center, Jerusalem, Israel); Hajime Nakamura (Department of Pediatrics, Kobe University Graduate School of Medicine, Kobe, Japan). REFERENCES 1 American Academy of Pediatrics. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004; 114: 297–316. 2 Bhutani VK, Johnson L, Sivieri EM. Predictive ability of a predischarge hour-specific serum bilirubin for subsequent significant hyperbilirubinemia in healthy term and near-term newborns. Pediatrics 1999; 103: 6–14. 3 Kaplan M, Muraca M, Hammerman C, Rubaltelli FF, Vilei MT, Vreman HJ et al. Imbalance between production and conjugation of bilirubin: A fundamental concept in the mechanism of neonatal jaundice. Pediatrics 2002; 110: e47–e51. 4 Maisels MJ, Pathak A, Nelson NM, Nathan DG, Smith CA. Endogenous production of carbon monoxide in normal and erythroblastotic newborn infants. J Clin Invest 1971; 50: 1–8.
Journal of Perinatology (2015), 1 – 4
5 Stevenson DK, Vreman HJ. Carbon monoxide and bilirubin production in neonates. Pediatrics 1997; 100: 252–254. 6 Stevenson DK, Bartoletti AL, Ostrander CR, Johnson JD. Pulmonary excretion of carbon monoxide in the human infant as an index of bilirubin production. II. Infants of diabetic mothers. J Pediatr 1979; 94: 956–958. 7 Stevenson DK, Bartoletti AL, Ostrander CR, Johnson JD. Pulmonary excretion of carbon monoxide in the human newborn infant as an index of bilirubin production: III. Measurement of pulmonary excretion of carbon monoxide after the first postnatal week in premature infants. Pediatrics 1979; 64: 598–600. 8 Stevenson DK, Ostrander CR, Hopper AO, Cohen RS, Johnson JD. Pulmonary excretion of carbon monoxide as an index of bilirubin production. IIa. Evidence for possible delayed clearance of bilirubin in infants of diabetic mothers. J Pediatr 1981; 98: 822–824. 9 Stevenson DK, Vreman HJ, Wong RJ, Contag CH. Carbon monoxide and bilirubin production in neonates. Semin Perinatol 2001; 25: 85–93. 10 Balaraman V, Pelke S, DiMauro S, Cheung S, Stevenson DK, Easa D. End-tidal carbon monoxide in newborn infants: observations during the 1st week of life. Biol Neonate 1995; 67: 182–185. 11 Stevenson DK, Fanaroff AA, Maisels MJ, Young BW, Wong RJ, Vreman HJ et al. Prediction of hyperbilirubinemia in near-term and term infants. Pediatrics 2001; 108: 31–39. 12 Bhutani VK, Stark AR, Lazzeroni LC, Poland R, Gourley GR, Kazmierczak S et al. Predischarge screening for severe neonatal hyperbilirubinemia identifies infants who need phototherapy. J Pediatr 2013; 162: 477–482 e471. 13 Vreman HJ, Wong RJ, Harmatz P, Fanaroff AA, Berman B, Stevenson DK. Validation of the natus CO-Stat end tidal breath analyzer in children and adults. J Clin Monit Comput 1999; 15: 421–427. 14 Beutler E, Gelbart T, Demina A. Racial variability in the UDPglucuronosyltransferase 1 (UGT1A1) promoter: a balanced polymorphism for regulation of bilirubin metabolism? Proc Natl Acad Sci USA 1998; 95: 8170–8174. 15 Bosma PJ, Chowdhury JR, Bakker C, Gantla S, de Boer A, Oostra BA et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyltransferase 1 in Gilbert's syndrome. N Engl J Med 1995; 333: 1171–1175. 16 Kaplan M, Renbaum P, Levy-Lahad E, Hammerman C, Lahad A, Beutler E. Gilbert syndrome and glucose-6-phosphate dehydrogenase deficiency: a dosedependent genetic interaction crucial to neonatal hyperbilirubinemia. Proc Natl Acad Sci USA 1997; 94: 12128–12132. 17 Shibahara S, Kitamuro T, Takahashi K. Heme degradation and human disease: diversity is the soul of life. Antioxid Redox Signal 2002; 4: 593–602. 18 Maisels MJ, Kring E. The contribution of hemolysis to early jaundice in normal newborns. Pediatrics 2006; 118: 276–279. 19 Dennery PA, Seidman DS, Stevenson DK. Neonatal hyperbilirubinemia. N Engl J Med 2001; 344: 581–590. 20 Newman TB, Xiong B, Gonzales VM, Escobar GJ. Prediction and prevention of extreme neonatal hyperbilirubinemia in a mature health maintenance organization. Arch Pediatr Adolesc Med 2000; 154: 1140–1147. 21 Keren R, Bhutani VK, Luan X, Nihtianova S, Cnaan A, Schwartz JS. Identifying newborns at risk of significant hyperbilirubinaemia: a comparison of two recommended approaches. Arch Dis Child 2005; 90: 415–421. 22 Tidmarsh GF, Wong RJ, Stevenson DK. End-tidal carbon monoxide and hemolysis. J Perinatol 2014; 34: 577–581. 23 Bhutani VK, Johnson LH, Jeffrey Maisels M, Newman TB, Phibbs C, Stark AR et al. Kernicterus: epidemiological strategies for its prevention through systems-based approaches. J Perinatol 2004; 24: 650–662. 24 Johnson LH, Bhutani VK, Brown AK. System-based approach to management of neonatal jaundice and prevention of kernicterus. J Pediatr 2002; 140: 396–403.
© 2015 Nature America, Inc.
