In July 2004, the Subcommittee on Hyperbilirubinemia of the American Academy of Pediatrics (AAP) published its clinical practice guideline on the management of hyperbilirubinemia in the newborn infant ≥35 weeks of gestation,1 and a similar guideline was published in 2007 by the Canadian Paediatric Society.2 Experience with implementation of the AAP guideline suggests that some areas require clarification. The 2004 AAP guideline also expressed hope that its implementation would “reduce the incidence of severe hyperbilirubinemia and bilirubin encephalopathy… .” We do not know how many practitioners are following the guideline, nor do we know the current incidence of bilirubin encephalopathy in the United States. We do know, however, that kernicterus is still occurring in the United States, Canada, and Western Europe.3–7 In 2002, the National Quality Forum suggested that kernicterus should be classified as a “serious reportable event,”8 sometimes termed a “never event,”9...
Skip Nav Destination
Article navigation
October 2009
Commentaries|
October 01 2009
Hyperbilirubinemia in the Newborn Infant ≥35 Weeks’ Gestation: An Update With Clarifications
M. Jeffrey Maisels, MB, BCh, DSc;
M. Jeffrey Maisels, MB, BCh, DSc
aDepartment of Pediatrics, Oakland University William Beaumont School of Medicine and Division of Neonatology, Beaumont Children's Hospital, Royal Oak, Michigan
Search for other works by this author on:
Vinod K. Bhutani, MD;
Vinod K. Bhutani, MD
bDepartment of Neonatal and Developmental Medicine, Lucile Salter Packard Children's Hospital, Stanford University, Palo Alto, California
Search for other works by this author on:
Debra Bogen, MD;
Debra Bogen, MD
cDivision of General Academic Pediatrics, Department of Pediatrics, University of Pittsburgh School of Medicine, Children's Hospital of Pittsburgh, Pittsburgh, Pennsylvania
Search for other works by this author on:
Thomas B. Newman, MD, MPH;
Thomas B. Newman, MD, MPH
dDepartment of Epidemiology and Biostatistics, Department of Pediatrics, University of California, San Francisco, California
Search for other works by this author on:
Ann R. Stark, MD;
Ann R. Stark, MD
eDepartment of Pediatrics and Section of Neonatology, Baylor College of Medicine and Texas Children's Hospital, Houston, Texas
Search for other works by this author on:
Jon F. Watchko, MD
Jon F. Watchko, MD
fDivision of Newborn Medicine, Department of Pediatrics, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania
Search for other works by this author on:
Address correspondence to M. Jeffrey Maisels, MB, BCh, DSc, Beaumont Children's Hospital, 3601 W. 13 Mile Rd, Royal Oak, MI 48073. E-mail: [email protected]
Pediatrics (2009) 124 (4): 1193–1198.
Article history
Accepted:
June 03 2009
Citation
M. Jeffrey Maisels, Vinod K. Bhutani, Debra Bogen, Thomas B. Newman, Ann R. Stark, Jon F. Watchko; Hyperbilirubinemia in the Newborn Infant ≥35 Weeks’ Gestation: An Update With Clarifications. Pediatrics October 2009; 124 (4): 1193–1198. 10.1542/peds.2009-0329
Download citation file:
Sign in
Don't already have an account? Register
Purchased this content as a guest? Enter your email address to restore access.
Please enter valid email address.
Pay-Per-View Access
$35.00
Citing articles via
Email alerts
Comments
Newborn jaundice nomograms: the wisdom of using imperfect but useful clinical tools.
Back in 1999, Bhutani described the first hour based serum bilirubin nomogram, [1] which opened a new way to predict and manage neonatal jaundice in the era of short hospital stay. This was undoubtedly the first clinical tool enough reliable and easy to use for this aim, in fact it has been widespread applied and incorporated into the AAP Guidelines. [2]
The development of accurate 2nd generation transcutaneous bilirubinometers provided the possibility to have repeated bilirubin measurements. Thus, Maisels [3] and our group [4] described two similar nomograms based on skin bilirubin in large newborn populations. We must also admit that the idea of developing bilirubin nomograms has been followed by other colleagues in different clinical contexts and nowadays several nomograms both on skin and serum bilirubin are available. [5,6,7] In this issue of Pediatrics, Varvarigou [8] published a last nomogram based on a Greek newborn population., while Fay and colleagues [9] criticised both this paper and the nomogram strategy to manage jaundice. They basically highlighted 3 problems:
1. Low worldwide applicability of nomograms, because curves are drawn on single centre populations. 2. Biased predictive ability of the nomograms, because of flawed methodology. 3. Uncertain clinical significance of the target that nomograms would predict (significant hyperbilirubinemia).
Does nomogram strategy have problems?
1. Actually, some basic characteristics of the Varvarigou¡¯s study [8] could affect its nomogram reliability, in the same way it can happen with our nomogram or Bhutani¡¯s or Maisels¡¯ one. Indeed, they are built on a single-centre population carrying typical traits which can be clearly different from other newborn cohorts. This is true not only for the basic jaundice risk factors (prevalence of caesarean section, breastfeeding, late preterm and positive direct Coombs¡¯ incidence), but also for other factors which may even be unknown (G6PD status, haemoglobinopathies, haematocrit, previous siblings receiving phototherapy etc¡). Moreover, genetic background may influence neonatal hyperbilirubinemia: we know that Black and East Asian infants are at lower and higher risk, respectively but we cannot reassume this influence simply with this dicotomic information. UDP-glucuronosyl transferase polymorphism and many other genes may play a complex role. [10,11] Nevertheless, many nomograms are now available and the baseline data of each study must not hinder us to study the risk factors in each local population and choose the nomogram which is best fitted for our babies. Best clinical practice suggest this for all type of study, since it is difficult to imagine a clinical trial which enrol a population representing completely all patients affected by the same disease, with all clinical differences in every medical centre worldwide.
2. Fay and colleagues are right when report some methodological biases [9] and it is well known that Bhutani¡¯s curve leads to false high sensitive estimate just because enrolled neonates were those more jaundiced at the clinicians¡¯ discretion. If this leads to a significant bias, then this is an ¡°in dubio pro reo¡± mistake. We must also bear in mind that, nomogram strategy is the first one which mathematically approaches the jaundice management with some degrees of reliability. In fact, the simple clinicians¡¯ discretion or the jaundice visual assessment alone must be considered even more misleading, while the universal follow- up (drawing blood to all infants at definite time-points) is time and money consuming, worrisome and even unethical in some cases if¡a certain degree of clinicians¡¯ discretion is not applied! Here the use of enough reliable tools, like jaundice risk factors evaluation and nomograms (especially if using non-invasive bilirubin measurement), may fill the gap between an ideal perfect screening method and the physician discretion. Of note, some available nomograms are less affected by the described bias; for example our skin bilirubin nomogram [4] was drawn up to 96 hours of life in inpatient neonates because of our late discharge policy. The study of Varvarigou [8] is theoretically more bias-free, because it has a delayed discharge policy and an universal follow-up between 96 and 144 hours. In fact, 1624 neonates were tested between 96 and 120 hour and 534 thereafter. Another critique is that available nomograms have generally been validated on the same populations used for drawing them. Nevertheless, we must bear in mind that since 1999, Bhutani¡¯s nomogram has been used by many other studies as a tool to predict significant hyperbilirubinemia and it has also been compared with a nomogram specifically built in a racial different newborn group. [12] Moreover, in 2005, we described an Italian serum bilirubin nomogram [13] which was drawn in an hospital and then validated in a completely different population in another centre, showing very high reliability.
3. Significant hyperbilirubinemia is a clinical target which has been greatly discussed [14] and the border between it and the physiologic jaundice may be very thin. Therefore, the target we would predict is actually uncertain, and, as Fay criticised, [9] we do not have a gold standard reference. Indeed, this is true because of the jeopardized pathophysiology of the neonatal jaundice. Nevertheless, we do believe that we cannot refrain from accepting this target, since kernicterus (which would be our ideal target to predict) is an extremely rare event and statistical power for its forecasting is lacking. A lot of factors may influence bilirubin neuro-toxicity albeit many other are still unknown. Nonetheless, we have some bilirubin levels which are considered enough harmful to be officially recommended to start phototherapy. [2] These levels are probably far from being neurotoxic and are coming from expert opinion instead from randomised clinical studies. However, good clinical practice force us to use them, in the absence of more powerful data: adherence to guidelines, in such a complex context, may lead, if any, to err in favour of the patient. Our next step should be to better understand the complex interaction between serum and skin bilirubin. They are strictly related but basically distinct variables. In fact, to use skin bilirubin for clinical decisions, we needed specific transcutaneous nomograms, since we would not have continued to draw skin bilirubin values over serum curves. Similarly, we now need to know the skin bilirubin value at which phototherapy should be started, with good degree of safety.
Best available options
We do believe that in more than 10 years the idea of evaluating jaundice using nomograms has been enough explored and validated. Indeed, false negatives exist and have been described: [15,16,17] it would be unrealistic to create a perfect diagnostic tool with 100% sensitivity and specificity. Nomograms reliability should be enhanced applying clinical wisdom and we recently reviewed some new evidences [18] that may help us:
1. Using the risk zone together with a careful evaluation of some risk factors improves the reliability, therefore clinical data should not be forgotten when drawing values on the curve. [19]
2. Even though the incidence of false negative is not exactly known it is likely that they are caused by the different distribution of risk factors amongst neonates enrolled for drawing nomograms. To avoid this each neonatal unit should build its own nomogram or use the curve more similar to the local population carefully considering the risk factors in that particular setting. [18]
3. Having more than one value, by using non-invasive measurements, allows bilirubin rate of rise calculation which is potentially the best tool to reduce false negative incidence. [7] In fact, rate of rise higher than the one of the initial percentile describes a ¡°crossing centiles behaviour¡±. This is due to the inbalance between bilirubin production and excretion [20] and identifies neonates at risk even though initially placed in a low risk zone. Preliminary data suggest that rate of rise > 0.1 mg/dl/h is associated with crossing centiles behaviour [7] and should be considered to plan follow-up independently from the risk zone. Universal jaundice screening programme has been criticised because it would cost 5.7$ million to prevent one kernicterus, without having a strong evidence of the programme effectiveness. [19] Since kernicterus is a rare and complex condition it would be hard to have a powerful demonstration supporting the screening programme. The alternative is similarly expensive (an universal follow-up programme, as suggested by Fay, [9] would not be realistic in a early discharge era [18]) and given the lifetime assistance of babies surviving kernicterus, the net balance is likely to be positive.
More and above this, we are here to avoid a ¡°never event¡± [21] which, in most cases, is preventable using our present knowledge: we must use the best tools we have, applying our clinical wisdom, and accept to err, if any, in favour of the safety.
REFERENCES [1] 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¨C14. [2] American Academy of Pediatrics. Sub-Committee on Hyperbilirubinemia. Management of Hyperbilirubinemia in the newborn infants 35 or more weeks of gestation. Pediatrics 2004;114: 297¨C316. [3] Maisels MJ, Kring E. Transcutaneous bilirubin levels in the first 96 hours in a normal newborn population of ¡Ý35 weeks¡¯ gestation. Pediatrics 2006; 117: 1169¨C73. [4] De Luca D, Romagnoli C, Tiberi E, Zuppa AA, Zecca E. Skin bilirubin nomogram for the first 96 h of life in a European normal healthy newborn population, obtained with multiwavelength transcutaneous bilirubinometry. Acta Paediatr 2008; 97: 146-150. [5] Sarici SU, Serdar MA, Korkmaz A, Oran O, Tekinalp G, Yurdakok M, et al. Incidence, course and prediction of hyperbilirubinemia in near term and term newborns. Pediatrics 2004; 113: 775-780. [6] Kaplan M, Hammerman C, Feldman R, Brisk R. Predischarge bilirubin screening in glucose-6-phosphate dehydrogenase-deficient neonates. Pediatrics 2000;105:533¨C7. [7] Romagnoli C, De Luca D, Zuppa AA, Parenti D, Latella C. Could early serum bilirubin measurement be useful in predicting non physiologic hyperbilirubinemia? Ital J Pediatr 2005;31:52-60. [8] Varvarigou A, fouzas S, Skylogianni E, Mantagou L, Bougioukou B, Mantagos S. Transcutaneous bilirubinnomogram for prediction of significant neonatal hyperbilirubinemia. Pediatrics 2009; 124: 1052-1059 [9] Fay DL, Schellhase KG, Suresh GK. Bilirubin screening for normal newborns: a critique of the hour-specific bilirubin nomogram. Pediatrics 2009, 124: 1203-5 [10] Beutler E, Gelbart T, Demina A. Racial variability in UDP- glucoronosyl transferase (UG1TA1) promoter: a balanced polymorphism for regulation of bilirubin metabolism?. Proc Natl Acad Sci USA 1998; 95: 8170-4. [11] Maruo Y, Nishizawa K, Sato H, Doida Y, Shimada M. Association of neonatal hyperbilirubinemia with bilirubin UDP-glucuronosyl transferase polymorphism. Pediatrics 1999; 103: 1224-7. [12] Engle WD, Lai S, Ahmad N, Manning MD, Jackson GL. An hour-specific nomogram for transcutaneous bilirubin values in term and late preterm Hispanic neonates. Am J Perinatol 2009; 26: 425-30. [13] Romagnoli C, De Luca D, Zuppa AA, Parenti D, Latella C. Could early serum bilirubin measurement be useful in predicting non physiologic hyperbilirubinemia ?. Ital J Pediatr 2005; 31: 52-60. Available at: javascript:articolo('/articoli/2005/vol-1-05/romagnoli.pdf' [14] Maisels MJ. What¡¯s in a name ? Physiologic and pathologic jaundice: the conundrum of defining normal bilirubin levels in the newborn. Pediatrics 2006;118:805-7. [15] Johnson LH, Bhutani VK, Brown AK. System-based approach to management of neonatal jaundice and prevention of kernicterus. J Pediatr 2002; 140: 396-403. [16] Maisels MJ, Newman TB, Watchko JF. Effect of predischarge bilirubin screening on subsequent hyperbilirubinemia. Pediatrics 2006; 118: 1796; [17] Slaughter J, Annibale D, Suresh GK. False negative results of predischarge neonatal bilirubin screening to predict severe hyperbilirubinemia: a need for caution. Eur J Pediatr 2009 in press [18] De Luca D, Carnielli VP, Paolillo P. Neonatal hyperbilirubinemia and early discharge from the maternity ward. Eur J Pediatr 2009; 168:1025-30. [19] Maisels MJ, Butani VK, Bogen D, Newman TB, Stark AR, Watchko JF. Hyperbilirubinemia in the newborn infant ¡Ý 35 weeks¡¯ gestation: an update with clarifications. Pediatrics 2009; 124: 1193-8 [20] Maisels JM, Kring E. The contribution of hemolysis to early jaundice in normal newborns. Pediatrics 2006; 118: 276-279. [21]Davidson L, Thilo EH. How to make kernicterus a ¡°never event¡±. Neoreviews 2003; 4: 308-314.
Conflict of Interest:
None declared