Rebound hyperbilirubinemia prediction: a simplified 2-variable model works well

  • Pediatrics

  • International Clinical Digest
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.

Takeaway

  • A simplified 2-variable model accurately predicts rebound hyperbilirubinemia in infants after inpatient phototherapy.
  • Model uses gestational age (GA), difference between starting/ending total serum bilirubin (TSB).

Why this matters

  • Authors say that for infants

Key results

  • 4.6% of infants had rebound hyperbilirubinemia.
  • Area under the curve for the 2-variable model was 0.876 (95% CI, 0.854-0.899) vs 0.881 (95% CI, 0.859-0.903) with 3 variables.
  • Rebound probability with phototherapy cessation at 2 mg/dL below starting threshold:
    • For infants ≥38 weeks’ GA: 2.5%.
    • For infants  
  • Score calculation: 15.5 (if GA
  • Equation (can be entered as function in spreadsheet) for predicted probability of rebound: 1/(1 + exp[–(0.097 * score – 2.84)]).
  • The models had a mean difference of 0 for predicting rebound.

Study design

  • Retrospective cohort study, 52 infants born 2012-2014 at ≥35 weeks of gestation, 17 Kaiser Permanente hospitals.
  • Infants had inpatient phototherapy before age 14 days.
  • Funding: None.

Limitations

  • Relatively small sample.
  • 15% had glucose-6-phosphate dehydrogenase deficiency.