Primary risk stratification for neonatal jaundice among term neonates using machine learning algorithm

BACKGROUND

Neonatal jaundice occurs in approximately 60% of term newborns. Although risk factors for neonatal jaundice have been studied, all the suggested strategies are based on various newborn tests for bilirubin levels. We aim to stratify neonates into risk groups for clinically significant neonatal jaundice using a combined data analysis approach, without serum bilirubin evaluation.

STUDY DESIGN

Term (gestational week 37-42) neonates born in a single medical center, 2005-2018 were identified. Anonymized data were analyzed using machine learning. Thresholds for stratification into risk groups were established. Associations were evaluated statistically using neonates with and without clinically significant neonatal jaundice from the study population.

RESULTS

A total of 147,667 consecutive term live neonates were included. The machine learning diagnostic ability to evaluate the risk for neonatal jaundice was 0.748; 95% CI 0.743-0.754 (AUC). The most important factors were (in order of importance) maternal blood type, maternal age, gestational age at delivery, estimated birth weight, parity, CBC at admission, and maternal blood pressure at admission. Neonates were then stratified by risk: 61% (n = 90,140) were classed as low-risk, 39% (n = 57,527) as higher-risk. Prevalence of jaundice was 4.14% in the full cohort, and 1.47% and 8.29% in the low- and high-risk cohorts, respectively; OR 6.06 (CI: 5.7-6.45) for neonatal jaundice in high-risk group.

CONCLUSION

A population tailored "first step" screening policy using machine learning model presents potential of neonatal jaundice risk stratification for term neonates. Future development and validation of this computational model are warranted.

Diagnostic criteria and contributors to Gilbert's syndrome

Hyperbilirubinemia is a well-known condition in the clinical setting; however, the causes of elevated serum bilirubin are diverse, as are the clinical ramifications of this condition. For example, diagnoses of individuals vary depending on whether they exhibit an unconjugated or conjugated hyperbilirubinemia. Diagnoses can include conditions of disordered bilirubin metabolism (Gilbert's, Crigler-Najjar, Rotor, or Dubin-Johnson syndromes) or an acquired disease, including alcoholic/non-alcoholic fatty liver disease, hepatotropic hepatitis, cirrhosis, or hepato-biliary malignancy. Assessment of bilirubin concentrations is typically conducted as part of routine liver function testing. Mildly elevated total bilirubin with normal serum activities of liver transaminases, biliary damage markers, and red blood cell counts, however, may indicate the presence of Gilbert's syndrome (GS), a benign condition that is present in ∼5-10% of the population. In this case, mildly elevated unconjugated bilirubin in GS is strongly associated with "reduced" prevalence of chronic diseases, particularly cardiovascular diseases (CVD) and type 2 diabetes mellitus (and associated risk factors), as well as CVD-related and all-cause mortality. These reports challenge the dogma that bilirubin is simply a potentially neurotoxic by-product of heme catabolism and emphasize the importance of understanding its potential beneficial physiologic and detrimental pathophysiologic effects, in order to appropriately consider bilirubin test results within the clinical laboratory setting. With this information, we hope to improve the understanding of disorders of bilirubin metabolism, emphasize the diagnostic importance of these conditions, and outline the potential impact GS may have on resistance to disease.

Correlation of UGT1A1 TATA-box polymorphism and jaundice in breastfed newborns-early presentation of Gilbert's syndrome

OBJECTIVE

The etiology of jaundice in otherwise healthy breastfed newborns that can present as early-onset exaggerated physiologic jaundice, or late breast milk jaundice (BMJ), is not yet entirely understood. This study tested the hypothesis that molecular marker for Gilbert's syndrome (GS), UGT1A1 TATA-box polymorphism, is associated with this disorders.

METHODS

We have investigated the UGT1A1 polymorphism frequency and its relation to severity of hyperbilirubinemia and jaundice duration among 220 exclusively breastfed term newborns; 57 of them with non-physiologic hyperbilirubinemia (NH), and 163 with BMJ, and in 187 healthy controls.

RESULTS

Significant differences in TA7/7 genotype frequency were established. The highest frequency was observed among the newborns with BMJ (42.0%), intermediate in the NH group (24.6%), while the controls had the lowest TA7/7 frequency (12.8%). Linear increase in TA7/7 frequency was observed depending on the duration of jaundice, peaking at 42.4% in newborns with the longest jaundice duration. Positive correlation between the serum bilirubin levels and the TATA-box length was established in all groups.

CONCLUSION

This study provides evidence that UGT1A1 TATA-box polymorphism is an important risk factor for developing jaundice in term breastfed newborns, presented as either early non-physiologic hyperbilirubinemia or breast milk jaundice. These results further support the original Odell's idea of neonatal jaundice as an early presentation of GS.

Implementation of a prolonged neonatal jaundice protocol supported by electronic software

Purpose

Several audits of clinical practice have confirmed that there is insignificant yield from screening of prolonged jaundice babies with a wide range of tests including thyroid function test (TFT) and full blood count (FBC). This paper aims to audit local practice based on a newly adopted rationalised protocol of selective screening with split bilirubin supplemented with other investigations based on clinical risk assessment.

Design/methodology/approach

The paper uses a retrospective audit of 50 babies referred for prolonged jaundice to a large district hospital in South West England. The authors also designed an easy‐to‐use, user‐friendly software application as a quick aid for implementation of the prolonged jaundice protocol.

Findings

The overall annual incidence of prolonged jaundice was 2.2 per cent of all live births. 32 patients (64 per cent) had only split biliburin test according to the protocol. In total 16 patients (32 per cent) had 18 tests requested unnecessarily including seven liver function test (LFT), eight FBC, one each of glucose, TFT and serum electrolytes. No patient with any serious liver disease was missed.

Practical implications

Appropriate recommendations for optimising the implementation of the current protocol based on the available scientific evidence were introduced.

Originality/value

After two years of implementation of the protocol, it has been found to be cost‐effective and safe. The electronic software provides a useful prompt in guiding the investigation of prolonged jaundice.