Abstract
Purpose:
Controversy exists regarding the spectrum of bilirubin-induced neurologic dysfunction (BIND) in neonates, and at what level of bilirubin an infant should be treated to reduce the chances of developing BIND. The goal of this study was to compare sweep visual evoked potential (sVEP) thresholds for 3 measures of vision: contrast sensitivity (CS), vernier acuity (VA), and grating acuity (GA), to an algorithm that predicts risk of BIND based on the hour-specific bilirubin Bhutani nomogram. We hypothesized that infants at higher risk on the algorithm would show worse acuity thresholds.
Methods:
90 full-term infants were recruited at birth from the well baby nursery. None had any clinical evidence for hemolysis or other risk factors for developing jaundice. Bilirubin levels were determined by measurements of transcutaneous bilirubin (TcB), total plasma bilirubin (TB), and “free” or unbound bilirubin (UB) levels. At 6 and 12 mos of age, VA, GA, and CS thresholds were measured by a masked observer using the sVEP technique. Neonatal history was tracked prospectively for predischarge TB, use of phototherapy, exposure to anesthesia (none), heart disease (excluded), and thyroid dysfunction. Infants were divided into 4 quartiles, based on their hour-specific TB level. Infants in each quartile were compared to their average acuity thresholds at 6 and 12 mos of age.
Results:
TB levels ranged from 0.5-21.6 mg/dL (median 7.9). CS and VA (p<0.05 for 6 and 12 mos) were significantly correlated to TB levels overall, but the effect was most pronounced in infants in the top quartile at both 6 and 12 mo of age; i.e., those who have the highest for BIND. GA was not affected by TB level.<br />
Conclusions:
A reduction in visual sensitivity occurs at 6 mos and persists to 12 mos of age, and correlates with the level of TB in the newborn period for CS and VA measures. Infants whose TB levels were in the highest (4th) quartile on the Bhutani nomogram were significantly more likely to have reductions in visual sensitivity compared to infants in the lower 3 quartiles. These findings suggest that a strategy for managing neonatal jaundice, and for potentially preventing BIND, at least in regards to the visual cortex and related functions.<br />