Technical Report: Diagnosis and Management of Hyperbilirubinemia in the Newborn Infant 35 or More Weeks of Gestation

In January of 2014, Dr Alex Kemper was approved by the AAP Board of Directors as the Chair of the Clinical Practice Guideline Subcommittee on Hyperbilirubinemia. Dr Kemper was charged with leading the update of the hyperbilirubinemia clinical practice guideline (CPG) focused on the diagnosis and management of hyperbilirubinemia in newborn infants born at ≥35 weeks’ gestation. The rationale for development of an updated guideline included the following:

1. New evidence that bilirubin neurotoxicity does not occur until concentrations well above the 2004 exchange transfusion thresholds justified raising the phototherapy treatment thresholds by a narrow range; and

2. Increasing evidence that phototherapy may have rare but serious late adverse effects.

In April 2014, the AAP Executive Committee approved the final subcommittee members for the formation of the Clinical Practice Guideline Subcommittee on Hyperbilirubinemia.

The subcommittee consisted of individuals with expertise in neonatal hyperbilirubinemia and late-preterm and term newborn care, including members of multiple relevant AAP committees. The subcommittee was chaired by Alex R. Kemper, MD, MPH, MS, FAAP, who appointed Thomas B. Newman, MD, MPH, FAAP, as vice chair in 2017. Jonathan L. Slaughter, MD, MPH, FAAP, a neonatologist, served as epidemiologist and methodologist. He developed the initial evidence-review questions and evidence tables for the potential use of intravenous immune globulin (IVIG) and/or exchange transfusion. Three librarians—Susi Miller, Teri Ballard, and Allison Erlinger—assisted with the primary literature search. Kymika Okechukwu, MPA, was the AAP staff representative for the project. All conflicts of interest were disclosed at the beginning of the process and updated throughout.

The subcommittee met face-to-face in August of 2014, May of 2017, and virtually in August of 2018. Conference call and E-mail correspondence were frequent and used to assess the quality of evidence, reach consensus, and develop the final clinical practice guideline.

The AAP Evidence-Based Clinical Practice Guidelines Development and Implementation Manual¹⁴  recommends that clinical practice guidelines be reviewed periodically and updated if “literature surveillance suggests that significant changes in clinical practice would be supported by strong evidence or monitoring of implementations suggests that the current guideline results in care that would be unnecessarily harmful to patients.” Since the original 2004 clinical practice guideline²  and the 2009 update with clarifications,³  new data have emerged regarding potential phototherapy-related harms (Table 1),^{12,13,15–17}  along with evidence that the phototherapy and exchange transfusion treatment thresholds from the previous 2004 clinical practice guideline could be raised to safely reduce overtreatment.^5–11 

In addition, neonatal hyperbilirubinemia experts have noted the need for additional guidance to separate routine phototherapy from “escalation of care,” including more aggressive phototherapy, intravenous hydration for dehydrated infants, and the potential use of IVIG and/or exchange transfusion. In recent years, multiple randomized controlled trials evaluated the effectiveness of IVIG to prevent exchange transfusion in infants with indirect hyperbilirubinemia from hemolysis precipitated by maternal-infant blood group incompatibility. Most of the subject areas within the 2022 guideline update were previously included in the 2004 clinical practice guideline²  and/or 2009 commentary.³  Accordingly, for most subjects, the committee reviewed the peer-reviewed literature for new, subsequent evidence since 2009. Complete systematic reviews were conducted and new evidence-tables created for two topics: phototherapy-related harm and the effectiveness of IVIG to prevent exchange transfusion in infants with indirect hyperbilirubinemia. For phototherapy-related harm, an emerging subject that was not previously covered, the committee reviewed all relevant literature without a date restriction. In addition, the committee reviewed all randomized controlled trials on the effectiveness of IVIG to prevent exchange transfusion in infants with indirect hyperbilirubinemia, to develop an evidence-table comparing recent placebo-controlled trials of infants with Rhesus (Rh) incompatibility to previous literature on IVIG treatment of infants with maternal-infant blood group incompatibility-mediated hemolysis.

The two systematic reviews were conducted in accordance with the “Preferred Reporting Items for Systematic Reviews and Meta-Analysis Guideline.”¹⁸  The epidemiologist developed questions to guide the literature search, evidence abstraction, and construction of the tables. The primary literature search was assisted by librarians Susi Miller, Teri Ballard, and Allison Erlinger.

The epidemiologist and librarians created a list of search terms and strategies.

Question 1: What are the adverse clinically detectable effects of phototherapy in newborns?

Population: Neonates born at ≥35 weeks’ gestation with indirect hyperbilirubinemia who are candidates for phototherapy.

Intervention: Treatment with phototherapy.

Comparator: Nontreatment.

Outcome: Increased risk of any adverse clinical outcome that is important to neonatal patients, their families, and/or the clinicians treating them.

The first search (conducted May 24, 2018; repeated March 5, 2021 and April 7, 2022) focused on medical diagnoses and adverse effects associated with phototherapy and was age-limited to newborn infants from birth to one month of age with the Medline limiter “newborn infant (birth to one month).” It was further limited using the Medical Subject Headings (MeSH) term: Phototherapy with subheading Adverse Effects. The search strategy was expanded to include phototherapy (keyword) AND – any of the following MeSH terms or keywords (newborn*; neonat*; adverse; asthma; Asthma[MeSH]; epilepsy; Epilepsy[MeSH]; cancer; Neoplasms[MeSH]; neuro*; seizure*; Seizure[MeSH]).

A second search (October 1, 2018; repeated March 5, 2021 and April 7, 2022) focused on the effects of phototherapy on family bonding and breastfeeding. The search strategy included searching Medline for articles indexed with either Phototherapy[MeSH term] OR Neonatal Jaundice[MeSH term with subheading Therapy] AND ALSO one of the following MeSH terms: Mothers/subheading psychology; Maternal Behavior; Parent-Child-Relations; Breast Feeding; breastmilk; Family Relations OR one of the following key word/phrases: “maternal-infant bond” or “mother-infant separation.” Our search strategy also including searching for keyword “phototherapy” AND one or more of the following keywords or phrases: “breastfeeding,” “bonding,” “separation,” “family dynamics,” or “family relationships.” A final search strategy including searching for articles indexed with both of the following: Object attachment[MeSH term] AND Phototherapy[MeSH term}. Case reports or series, nonsystematic review articles, and manuscripts focused on preterm infants born at <35 weeks’ gestation were excluded.

Question 2: How effective is IVIG for preventing exchange transfusion in infants with indirect hyperbilirubinemia?

Population: Neonates born at ≥35 weeks’ gestation with isoimmunization and hyperbilirubinemia who are at risk for exchange transfusion.

Intervention: Treatment with IVIG.

Comparator: Nontreatment with IVIG.

Outcome: Need for exchange transfusion.

The search, initially performed on May 24, 2018 and repeated March 5, 2021 and April 7, 2022, included MeSH term Hyperbilirubinemia or keyword hyperbilirubinemia AND – any of the following MeSH terms or keywords (Immunoglobulins, Intravenous[MeSH]; IVIG; intravenous immunoglobulin*; immunoglobulin*). Commentaries, editorials, and letters were excluded.

After the initial search produced multiple randomized trials, the article inclusion criteria were subsequently limited to randomized controlled trials, quasi-randomized trials, and systematic reviews that included randomized trials.

References obtained from Medline via the search process were initially selected on the basis of the article title and abstract and the prewritten inclusion criteria for each question. Each reference was reviewed by 2 reviewers (JLS and ARK), and a third reviewer (TBN) was the deciding vote in instances of disagreement. The first 2 reviewers subsequently reviewed the full texts of the included articles.

Question 1: What are the adverse clinically detectable effects of phototherapy in newborns?

The search produced 45 manuscripts. Eighteen were excluded: 12 did not report clinically relevant phototherapy-related outcomes, 2 compared phototherapy modalities rather than outcomes, 1 focused on preterm infants born at <35 weeks’ gestation, 1 was a case report, 1 a case series, and 1 a nonsystematic review article. A review of the selected manuscripts yielded 14 additional publications which were included. An evidence table (Table 1) details the study design of and evidence derived from each of the 38 included manuscripts forquestion 1.

Summary: Phototherapy is associated with a significant yet low overall risk of potential harm. Childhood seizures are one of the most serious, but infrequent, conditions (adjusted 10-year excess risk of 2.4; 95% confidence interval [CI]: 0.6, 4.1 per 1000 phototherapy-treated infants) associated with phototherapy.¹²  Although the topic is understudied, there is some evidence that phototherapy may limit familial bonding with the infant. Therefore, treatment thresholds must attempt to balance the risk of adverse effects of phototherapy with its possible benefit at reducing the risk of total serum bilirubin concentrations at which exchange transfusion is recommended and/or neurotoxicity has been found.

Question 2: How effective is IVIG for preventing exchange transfusion in infants with indirect hyperbilirubinemia?

The search produced 13 manuscripts. Two were excluded because they were not randomized trials, quasi-randomized trials, or systematic reviews inclusive of randomized trials. An evidence table (Table 2) details the study design of and evidence derived from each of the 11 included manuscripts for question 2.

Summary: IVIG may not be effective in preventing exchange transfusion for infants with severe Rh disease. Larger, randomized, placebo-controlled trials are needed to confirm this finding. The effect of IVIG for non-Rh-mediated hemolysis including ABO incompatibility is understudied. There is a large degree of heterogeneity between available trials, which are all of low to moderate quality. Thus, there is a need for well-designed placebo-controlled trials. There were no IVIG-related safety concerns reported in the included randomized trials. However, observational investigations have reported associations between IVIG and necrotizing enterocolitis (NEC). Figueras-Aloy et al conducted a retrospective cohort investigation in Barcelona, Spain, and found an association between IVIG and risk of NEC (odds ratio [OR]: 31.66; 95% CI: 3.25–308.57) but no increase in mortality.¹⁹  The authors attempted to adjust for severity of illness using propensity scoring PS, but it does not appear that all of the included variables (eg, peak bilirubin concentration or exchange transfusion) were measured before the receipt of IVIG, as should be done when using propensity scoring to mimic a randomized trial. Although care practices mostly seemed similar to the United States, they did administer phenobarbital to infants receiving phototherapy. A meta-analysis that included the Figueras-Aloy et al investigation specifically evaluated the association between IVIG and NEC and reported a higher incidence of NEC after IVIG infusion for hyperbilirubinemia (OR: 4.53; 95% CI, 2.34–8.79) with no increase in mortality.²⁰  However, the study was limited by inclusion of multiple nonrandomized studies, an unclear definition of NEC grade, a large degree of heterogeneity (eg, proportion of small for gestational age infants) between available studies, and an incomplete description of the evaluated studies including the baseline severity of illness in infants who were treated with IVIG, as compared to untreated infants with hyperbilirubinemia.¹⁹  Despite these limitations, the signal for IVIG-related harm from increased NEC must be considered, because observational studies are often better powered than randomized trials to pick up rare, serious adverse drug events. Given an unclear but potential benefit from IVIG in preventing exchange transfusion and limited evidence of an increased NEC risk, the risks and benefits of IVIG must be carefully considered. The 2022 guideline notes that clinicians may consider the administration of IVIG to infants with isoimmune hemolytic disease (ie, positive direct antiglobulin test [DAT]) who have not initially responded to photherapy with a reduction in total bilirubin only in circumstances when the TSB is rising despite intensive phototherapy or within 2 to 3 mg/dL of the exchange level and there is concern that a timely exchange transfusion will be difficult.¹ 

All first drafts of recommendations were created by members of the subcommittee that were assigned as content experts and authors for each subsection of the guideline. In many instances, these recommendations were closely adapted from similar recommendation statements in the 2004 guideline and/or the 2009 commentary by reviewing and incorporating any new evidence on each previous recommendation topic, when available. The original drafts of recommendations for the 2 questions, as derived from the systematic reviews and evidence tables, were written by the epidemiologist.

All members of the subcommittee then reviewed all recommendation drafts and offered amendments before a final acceptance of all recommendations by vote of the entire subcommittee. The derived language for each recommendation reflected the strength of graded evidence.

The identified evidence for each recommendation within the clinical practice guideline was appraised and summarized per AAP policy.¹⁴  Then a level of evidence, from strongest to weakest, was assigned (Fig 1).¹⁴  This information is provided in Supplemental Appendix B of the clinical practice guideline.¹ 

A systematic approach combining the best available evidence and expert opinion, when necessary, was used to generate an updated guideline for the diagnosis and management of hyperbilirubinemia in neonates ≥35 weeks’ gestation. The committee’s approach was to improve on the expert-driven 2004 AAP guideline²  and 2009 follow-up commentary³  through the incorporation of new evidence. The committee has developed new phototherapy and exchange transfusion treatment recommendations that should reduce the numbers of infants unnecessarily treated with phototherapy while at the same time ensuring that infants at higher risk for bilirubin-associated neurotoxicity, on the basis of serum bilirubin concentrations, gestational age, and neurotoxicity risk factors, will receive proper treatment to prevent hyperbilirubinemia-associated harm.¹ 

We thank librarians Susi Miller, Teri Ballard, and Allison Erlinger at Nationwide Children’s Hospital, Columbus, Ohio.