Guidance for the Primary Care Provider in Identifying Infants With Biliary Atresia by 2–4 Weeks of Life: Clinical Report

A critical period in biliary atresia (BA) is the time between birth and treatment with the Kasai portoenterostomy. Ideally, infants are treated before 30 to 45 days of life to achieve the best outcomes.^1–3 In practice, however, infants are treated on average after 60 days of life in the United States.^4–7 Delays arise because BA can be asymptomatic initially. When symptoms do develop, the 1 in 8000 to 18 000 affected newborn infants each year are difficult to distinguish from the 65% of healthy newborn infants with benign causes of jaundice.^4,8 In part because of these delays, BA outcomes remain poor, with BA accounting for 60% of liver transplants in infants younger than 1 year and 30% of all liver transplants in children.⁹ 

Pediatric primary care providers (PCPs) have a unique opportunity to improve outcomes by identifying BA during the first well-child visits. This clinical report provides a strategy to be used between 2 and 4 weeks of life at the “By 1 month” well-child visit in the Bright Futures/American Academy of Pediatrics (AAP) “Recommendations for Preventive Pediatric Health Care” (also called the periodicity schedule).¹⁰ The strategy involves examining every infant’s eye color, stool color, and prior laboratory results to determine whether drawing a direct or conjugated bilirubin level is warranted (Figure 1).

Jaundice beyond 2 to 3 weeks of life warrants consideration of testing for direct or conjugated bilirubin levels, according to joint recommendations from North American and European pediatric gastroenterology societies (see “Billing and Scheduling Considerations” below for a discussion on timing).⁷ Jaundice is the key physical examination finding, because other signs—such as hepatomegaly and splenomegaly—may be absent in early stages of BA.

If the answer to Step 1 is “No,” the strategy proceeds to Step 2 because an infant could still have BA. For example:

• Jaundice may not be present yet. In early stages, infants with BA may not have visible color changes in the eyes or skin. Visible signs can be lacking even though serum direct or conjugated bilirubin levels are elevated.¹¹ 

• Jaundice may not be noticed. Certain infants may be vulnerable to a less careful examination because of an assumption that skin assessment is unreliable in infants who have skin with more melanin (importantly, at least one study challenges this assumption; see “Future Directions” section below). To help prevent missed cases, this report emphasizes the importance of assessing eye color, rather than just skin color, for jaundice.

Pale, gray, or white stools at any time warrant consideration of testing for direct or conjugated bilirubin levels. Lighter stools occur in BA and other liver diseases when conjugated bilirubin in bile fails to reach the intestine and color the stools.

If the answer to Step 2 is “No,” the strategy proceeds to Step 3 because an infant could still have BA. For example:

• The stools have not changed color yet. The transition to pale, gray, or white stools occurs over time in BA. Approximately 77% of infants with BA have pale, gray, or white stools at 30 days of life.¹² This proportion increases to 84% and 97% at 45 and 60 days of life, respectively.¹³ 

• A change in stool color may not be noticed. To avoid this, stool color can be assessed in multiple ways, including (i) asking parents about stool color; (ii) inspecting stools at the clinic visit; (iii) reviewing pictures of stool color; (iv) comparing stool color to the “stool color card;” and/or (v) assessing stool color with a phone app such as PopòApp (Figure 2 shows images from a sample stool color card).^14,15 

A previously “high” direct or conjugated bilirubin level warrants consideration of repeat testing. To determine whether previous values were “high,” prior laboratory results must be reviewed from the newborn inpatient unit and/or outpatient setting. Many infants have total bilirubin levels checked in accordance with AAP recommendations to assess risk for bilirubin encephalopathy.¹⁶ Alongside total bilirubin, levels of direct or conjugated bilirubin are often measured.

The most important result is the initial direct or conjugated bilirubin level, which will be “high” in BA starting at birth.^11,17,18 “High” is defined as exceeding the laboratory’s derived reference range, even if only by 0.1 mg/dL (Figure 1, footnote *). In the period before 2 weeks of life, “high” is not defined by exceeding a fixed cut-off or exceeding a bilirubin ratio (see  Appendix for further discussion of interpreting initial values).^11,17 

If answers to questions in Step 3A or 3B are “No,” skin and stool color should continue to be monitored because an infant could still have BA. For example:

• A previous direct or conjugated bilirubin level may not have been obtained. This can occur if risk for bilirubin encephalopathy was assessed by (i) total bilirubin levels alone, without also measuring direct or conjugated bilirubin levels; or (ii) transcutaneous measurements, which provide no information about direct or conjugated bilirubin levels.

• A direct or conjugated bilirubin level is compared to the wrong reference range. For interpreting levels drawn before 2 weeks of life, using the correct reference range is critical. Reference ranges for direct or conjugated bilirubin are laboratory-specific and can vary widely across hospitals (in one multicenter study, the derived reference ranges for participating hospitals varied from 0.0–0.2 mg/dL to 0.0–0.7 mg/dL).¹⁹ As a result, in rare instances, a level may be high but incorrectly reported as normal if the laboratory used a borrowed—rather than derived—reference range (see  Appendix for further discussion of deriving reference ranges).²⁰ 

Based on the strategy in Figure 1, jaundiced eyes, pale stools, or a high initial laboratory result warrants consideration of testing for direct or conjugated bilirubin levels. However, a blood draw can possibly be avoided if levels checked earlier in the newborn inpatient unit or outpatient setting are available (Figure 1, footnote **). Infants for whom a blood test can be avoided have 1 of the following:

• Any prior direct or conjugated bilirubin level that was normal. In this instance, normal is defined as within the laboratory-derived reference range (normal is not defined with fixed cut-offs or bilirubin ratios in this early time period). Infants with normal levels are unlikely to have BA. This is because direct or conjugated bilirubin levels are above the reference range in BA starting at birth.^11,17 

• Prior direct or conjugated bilirubin levels that were all abnormal but equivalent or decreasing over time. In this instance, equivalent or decreasing is defined as both (i) less than or equal to the initial level; and (ii) <1 mg/dL. Infants with levels decreasing over time are unlikely to have BA. This is because direct or conjugated bilirubin levels increase over the first few weeks of life in BA before plateauing.²¹ 

Upon drawing a direct or conjugated bilirubin level, a value ≥1 mg/dL warrants urgent consultation with a pediatric gastroenterologist. The pediatric gastroenterologist will interpret the values and may recommend the following to evaluate for BA:

• Perform additional testing. The pediatric gastroenterologist may ask only for the direct or conjugated bilirubin to be re-checked, to confirm the initial abnormal result. In other cases, the pediatric gastroenterologist may ask for more tests such as liver enzymes or an abdominal ultrasound. However, ordering liver enzymes or an abdominal ultrasound before consulting a pediatric gastroenterologist is not recommended, because normal results could be falsely reassuring and delay referrals.^22,23 

• Refer to gastroenterology. The pediatric gastroenterologist may prefer to see the patient in an outpatient setting before ordering further testing. An urgent referral should be made, and the infant’s family should understand the importance of not missing the appointment. Ideally the infant is seen by pediatric gastroenterology within 5 days.

• Send to emergency department. In some cases, the pediatric gastroenterologist may elect to expedite the evaluation with an inpatient admission. To do this, the pediatric gastroenterologist may ask that the infant be sent to the emergency department to begin the process.

If the direct or conjugated bilirubin level is <1 mg/dL but above the reference interval at 2 to 4 weeks of life, urgent consultation with a pediatric gastroenterologist is not needed. These infants are unlikely to have BA. This is because direct or conjugated bilirubin levels in BA increase to ≥1 mg/dL over the first few weeks of life.^17,21 Possible next steps include monitoring clinically, and no further testing is needed unless signs such as jaundice or pale stools develop.

The strategy outlined in this report may require separately identifiable work in addition to what is performed during the routine 2- to 4-week well-child visit. For example, using this strategy may involve more time and/or more complex decision making. In these instances, separate evaluation and management services may be able to be reported, in accordance with Current Procedural Terminology (CPT) guidelines and payer policies.

From a scheduling perspective, the Bright Futures/AAP periodicity schedule allows PCPs to decide the timing of this well-child visit (in contrast, timings for the “3–5 day” and “2-month” well-child visits are specified).¹⁰ Scheduling the visit at 2 weeks will help identify infants with BA earlier and help ensure that treatment with the Kasai portoenterostomy occurs before 30 days of life. However, an earlier visit may result in more infants being tested, because in one study 34% of healthy predominantly breastfed infants still appeared jaundiced at 3 weeks of life.²⁴ In contrast, scheduling the visit at 4 weeks may result in fewer healthy infants tested but may miss an opportunity for earlier treatment.

The first 2 patient encounters in the Bright Futures/AAP periodicity schedule—the “Newborn” and “3- to 5-day” encounters—can also be used to help identify infants with BA (Table 1). The “Newborn” encounter occurs before hospital discharge. This visit can be used to review direct or conjugated bilirubin levels and to consult a pediatric gastroenterologist in special cases, such as when levels are already ≥1 mg/dL and increasing. The “3–5 day” encounter occurs shortly after discharge. It can be used to (i) remind parents to look for pale stools; (ii) review prior direct or conjugated bilirubin levels; and (iii) schedule a date for the “By 1 month” well-child visit. At the “3–5 day” encounter, a pediatric gastroenterologist can also be consulted in special cases, such as when levels are already ≥1 mg/dL and increasing.

In addition, health care providers in the newborn inpatient unit can help by:

• Ensuring their laboratories derive reference ranges. Each hospital should establish its own reference range for direct or conjugated bilirubin levels (see  Appendix for further discussion of deriving reference ranges). Accurate reference ranges will ensure that all high direct or conjugated bilirubin levels are recognized.

• Consider adding direct or conjugated bilirubin testing when serum total bilirubin testing is performed. As mentioned earlier, many centers measure at least 1 serum total bilirubin level via heel stick or venipuncture to assess risk for bilirubin encephalopathy. Direct or conjugated bilirubin levels can be measured from the same heel stick or venipuncture sample, without needing an additional blood draw (future studies are needed to determine whether cord blood can also be used to identify BA).

• Communicating with outpatient PCPs. In the outpatient setting, there is limited time to search through discharge summaries or call birth hospitals to obtain direct or conjugated bilirubin results. Newborn inpatient providers could help by clearly documenting high levels and/or directly informing PCPs about abnormal results.

A future goal is to develop a universal screening strategy for BA, which would address the AAP goal of eliminating health disparities arising from current practice. Multiple studies have reported that Hispanic and non-Hispanic Black infants with BA are referred later than non-Hispanic white infants.^5,22,25,26 One potential reason is that some families miss well-child visits after infants are discharged from the hospital for reasons related to social determinants of health. Another potential reason is that identification of some infants is hindered by an implicit bias, which assumes that visually assessing skin color in individuals who have skin with more melanin is unreliable. When formally examined in one study, jaundice was recognized equally in Black and white infants.²⁷ 

There are at least 2 promising options for a universal screening strategy. The first option is the stool color card program, which screens infants for pale stools. The stool color card program has been implemented successfully in several locations, including regions in Japan, Taiwan, provinces in Canada, and Switzerland.^{12,13,28–31} The second option is newborn direct or conjugated bilirubin screening before hospital discharge.^32–35 The strategy has been implemented in parts of the United States.^19,36–38 Both strategies have been associated with earlier ages of treatment with the Kasai portoenterostomy, and future implementation studies will examine how well they can be integrated into standard pre-existing newborn management protocols.

PCPs can identify infants who may have BA at 2 to 4 weeks of life, using a strategy at the “By 1 month” well-child visit in the Bright Futures/AAP periodicity schedule. The strategy involves measuring a direct or conjugated bilirubin level based on an infant’s eye color, stool color, and prior laboratory results. By identifying infants earlier, PCPs have a unique opportunity to improve outcomes and help reduce the tremendous liver transplant burden of BA.

The body excretes heme by converting it to bilirubin, which is measured in 3 forms:

• Unconjugated bilirubin. Unconjugated bilirubin is made in the spleen by removing iron from heme molecules. Unconjugated bilirubin is water insoluble and cannot be excreted without further modification. Unconjugated bilirubin levels increase in the neonatal period during the process of normal fetal red blood cell/fetal hemoglobin turnover. Excessively high unconjugated bilirubin levels are thought to cause bilirubin encephalopathy, which can be prevented by following AAP phototherapy guidelines.¹⁶ 

• Conjugated bilirubin. Conjugated bilirubin is made in the liver by modifying unconjugated bilirubin to become water soluble. Conjugated bilirubin then dissolves in bile and flows through bile ducts into the intestine for excretion (conjugated bilirubin derivatives give stool its characteristic color). In BA, conjugated bilirubin cannot leave the liver because the bile ducts are obstructed. Instead, conjugated bilirubin accumulates in blood and can be detected at birth.

• Delta bilirubin. Delta bilirubin is made when conjugated bilirubin binds to albumin. Delta bilirubin only forms in pathological situations, when conjugated bilirubin accumulates chronically in the blood. Clearance of delta bilirubin follows the half-life of albumin.³⁹ Delta bilirubin levels are measured with the direct, but not conjugated, assay (see below).

Clinical questions related to measuring these bilirubin forms include:

Direct and conjugated bilirubin levels are used interchangeably but are not exactly the same (Appendix Table 1).^40,41 Both direct and conjugated assays measure conjugated bilirubin. In addition, the direct assay measures delta bilirubin as well as small amounts of unconjugated bilirubin. As a result, the direct assay always produces slightly higher results than the conjugated assay. The direct assay is used by more laboratories in the United States, although the direct assay results are sometimes incorrectly reported as “conjugated bilirubin.”

The first direct or conjugated bilirubin levels in BA—although elevated—may be only slightly higher than the reference range and still below 1 mg/dL.^11,17,18 For this reason, levels measured before 2 weeks of life are interpreted with laboratory-specific reference ranges (see below for a discussion on how reference ranges are derived). After 2 weeks of life, levels in BA will have increased sufficiently and can be interpreted using the fixed 1 mg/dL cut-off.

Using bilirubin ratios in the first days of life will miss infants with BA.^11,17 Bilirubin ratios are calculated by dividing the direct or conjugated level by the total level. Bilirubin ratios are often used to evaluate infants with elevated direct or conjugated bilirubin levels, to distinguish between non-liver causes (such as increased red blood cell destruction, ratio <0.2) versus liver causes (ratio ≥0.2). However, in one study, bilirubin ratios were ≥0.2 before 48 hours of life in only 21% of newborn infants later diagnosed with BA.¹¹ Bilirubin ratios may be more useful for identifying BA when used at later time periods.

Each hospital laboratory should derive its own direct bilirubin or conjugated bilirubin reference range. This is especially important for the direct assay, which can vary with slight changes in pH, temperature, and/or reaction times that occur across laboratories.⁴⁰ To derive reference ranges, laboratories can follow standard procedures of measuring serum from 120 newborn infants and designating the middle 95% of values (2.5^th–97.5^th percentiles) as normal.⁴² 

Sanjiv Harpavat, MD, PhD, FAAP

Susan Aucott, MD, FAAP

Saul Karpen, MD, PhD, FAAP

Benjamin Shneider, MD, FAAP

Kasper Wang, MD, FAAP

Mitchell B. Cohen, MD, FAAP, Chairperson

David Brumbaugh, MD, FAAP

Jennifer L. Dotson, MD, FAAP

Sanjiv Harpavat, MD, PhD, FAAP

Jenifer R. Lightdale, MD, FAAP, Immediate Past Chairperson

Daniel Mallon, MD, FAAP

Maria M. Oliva-Hemker, MD, FAAP

Debra L. Burrowes, MHA

Eric C. Eichenwald, MD, FAAP, Chairperson

Charletta Guillory, MD, FAAP

Ivan Leslie Hand, MD, FAAP

Mark L. Hudak, MD, FAAP

David Alan Kaufman, MD, FAAP

Camilia Rivera Martin, MD, FAAP

Margaret Graham Kemper Parker, MD, FAAP

Arun Kumar Pramanik, MD, FAAP

Kelly Cant Wade, MD, FAAP

Jim Couto, MA

Andrew Davidoff, MD, FAAP, Chairperson

Elizabeth Beierle, MD, FAAP

Gail E. Besner, MD, FAAP, Immediate Past Chairperson

Marybeth Browne, MD, FAAP

Cynthia D. Downard, MD, FAAP

Kenneth Gow, MD, FAAP

Saleem Islam, MD, FAAP

Danielle Walsh, MD, FAAP

Vivian Thorne

Christina Jung, MD, FAAP

Christoph U. Lehmann, MD, FAAP

Kymika Okechukwu

The authors would like to acknowledge Debra L. Burrowes, staff liaison to the AAP Section on Gastroenterology, Hepatology and Nutrition, for providing instrumental guidance through the permissions, drafting, and review process. The clinical report represents one of Ms. Burrowes’ last projects in her 17 year tenure with the AAP.

AAP

American Academy of Pediatrics

BA

biliary atresia

PCP

primary care provider