ABSTRACT
The prevalence of kernicterus in our neonatal intensive care unit (NICU) decreased
from more than 2/1000 live births in 1980 to none in 1984. To clarify predictors of
kernicterus, we examined the medical records of infants born during that time who
died between 2 and 28 days of age. Infants were divided into three groups: those with
kernicterus, all born before June 10, 1982 (n = 29); contemporaneous controls (n =
28); and infants born after June 10, 1982 (n = 32). Benzyl alcohol was not used after
June 10, 1982. Kernicteric infants were more likely than contemporaneous controls
to have seizures (p < 0.001). Indices of pulmonary disease, arterial partial pressure
of carbon dioxide and end-expiratory pressure on the ventilator were higher after
1982. However, exposure to benzyl alcohol was not different in kernicteric infants
and contemporaneous controls, suggesting that benzyl alcohol exposure was not the
explanation for the decrease in the prevalence of kernicterus in our NICU.
