Why Didn’t Screening for Biliary Atresia Improve Outcome In This Study?

Posted on by

ZJ Kastenberg et al J Pediatr 2023; 257: 113339. Fractionated Bilirubin Among 252 892 Utah Newborns with and Without Biliary Atresia: A 15-year Historical Birth Cohort Study

This retrospective study (2005-2019) used an administrative data from a large integrated healthcare network in Utah to identify newborns with abnormal fractionated bilirubins. since 2005, all newborns at this healthcare system had a fractionated bilirubin measured.

Key findings:

  • There were 252 892 newborns with fractionated bilirubin assessed, including 26 of those subsequently confirmed to have biliary atresia (BA).
  • Conjugated or direct bilirubin was elevated in all 26 infants with BA and an additional 3246 newborns (1.3%) without BA. The lowest direct bilirubin in the BA group was 0.6, just above cutoff value of 0.5 mg/dL. The conjugated bilirubin cutoff value wa 0.2 mg/dL.
  • The 15-year crude birth prevalence of BA was 0.68 per 10,000 births in this cohort
  • Median time to Kasai HPE was 69.5 days

This study found that all infants with BA have elevated conjugated or direct bilirubin at birth. The authors estimate that a healthcare system with about 30,000 deliveries per year, would have between 450 and 630 newborns that would require a second screen. In this group, 96 per 100,000 screened newborns (~1 in 1000) will have a positive second screen at 2-week well check and need further evaluation.

What I don’t understand about this paper is how the authors omit a discussion of the age of Kasai HPE. How is it that all of these newborns received a fractionated bilirubin and the age of Kasai HPE is not improved compared to other U.S locations that have not implemented universal testing? (Previous data from 15 ChiLDReN sites indicate age of Kasai HPE 65-70 days and unchanged over past 30 years, see blog post: Online Aspen Webinar (Part 5) -Biliary Atresia Diagnosis and Screening).

The authors note that implementation of screening could be improved with newer tests (eg. MMP-7) but that more analysis is needed to determine if screening is cost-effective and avoids harm (eg. subjecting healthy newborns to invasive testing).

My take: If universal screening is implemented, it is imperative to show that it helps and to set up the needed infrastructure to arrange appropriate followup. The first surrogate marker of this effort would be improving the age of surgical intervention.

As an aside, I find the new page numbering by The Journal of Pediatrics to be quite annoying. When you are going through the hard copy of the Journal, it is more difficult to find the articles that are listed in the table of contents because the articles are not ordered by lowest to highest numbered page. Each article is given a single page number like 113339 in this article but inside the article it is numbered page 1–10 (or however long the pages). The article prior could be a lower or higher page number. Until hard copies are eliminated, it would be an improvement if the articles could at least be ordered such that the next article would not have a lower page reference than the preceding article.

Related blog posts:

Castle Hill Park, Nice France