University of Michigan
Department of Pediatrics
Evidence-Based Pediatrics Web Site

Question

• A healthy term African-American male was found to be jaundice by visual inspection on the third day of life. Serum bilirubin levels obtained prior to discharge was found to be 18.5. Repeat bilirubin level the following day was 22.5 resulting in the patient being re-admitted to the hospital for phototherapy. Is there a reliable screening test which would have enabled the physicians to predict if this infant was at risk for developing severe hyperbilirubinemia and its associated complications prior to hospital discharge.
  

Clinical Bottom Lines

1. Neonatal jaundice is one of the most common diagnoses in the newborn nursery. Most cases of jaundice typically have a benign course. However, given the recent increase in the number of cases of kernicterus, neonatal jaundice continues to be an important public health issue.
   
2. Early recognition of infants at risk for severe hyperbilirubinemia can facilitate closer follow-up, early intervention and therapy that can prevent negative outcomes associated with severe hyperbilirubinemia.
   
3. The cost of obtaining a serum bilirubin level is $22.
   
4. Assuming physical exam does not pick up any baby with significant jaundice, the number needed to screen to detect babies with postdischarge bilirubin in the high risk zone is equal to 25.
   
5. This study does not look at differences in detection of significant hyperbilirubinemia in screened and unscreened children.

Summary of Key Evidence

1. Using data from this retrospective cohort study, a percentile-based bilirubin nomogram was constructed using hour specific predischarge and post discharge total serum bilirubin levels of 2840 healthy newborns at Pennsylvania Hospital.
   
2. Newborn infants were included if they were tem or near term greater than 2000g for greater than 36 weeks GA or BW greater than 2500g for greater than 35 weeks GA and were included in the hospital supervised follow-up program.
   
3. Infants were excluded if they were admitted to the NICU, or had a positive direct Coombs test. Newborns who required phototherapy <60 hrs of life were also excluded from the predictive nomogram.
   
4. Newborns were stratified to high risk (greater than 95% for age in hours), high-intermediate (between 75-95% for age in hours), low-intermediate (between 40-75% for age in hours) or low risk (less than 40% for age in hours) zones based on their hour specific TSB level.
   
5. High risk was >8 at 24 hours, >14 at 48 hours, >17 at 72 hours.
   
6. Study patients with predischarge bilirubin levels in the high risk zones were more likely to have post discharge bilirubin levels in the high risk zones. (LR=14.08)
   
7. Newborns with predischarge serum bilirubin levels in the intermediate risk zone were more likely to move into the high risk zone post discharge if they were in the higher intermediate group verses the lower intermediate group. (LR 3.2, and 0.48 respectively)
   
8. Study patients with predischarge bilirubin levels in the low risk zones were not likely to move into the high risk zone post discharge. (LR= 0)
   
9. Based on the nomogram's predictive ability, newborns with predischarge TSB levels in high and high intermediate risk zones could be followed closely post discharge to facilitate early targeted interventions.
   
10. Newborns in the low-risk zone could safely be managed by routine pediatrician follow-up with visual inspection.
    

Additional Comments

• Though one of the most common diagnoses in newborns, little evidence exists to support evidence-based approach to management.
  
• Study criteria excluded a segment of newborns that are particular risk for development of severe hyperbilirubinemia.
  
• What are the true advantages of this test over visual assessment? Is there a greater segment of patients that one would or would not catch?
  

Citation

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; 103:6-14.

CAT Author: Faye Holder MD

CAT Appraisers: Katherine Layton MD

Date appraised: January 11, 2002