Benzyl alcohol toxicity in a neonatal intensive care unit. Incidence, symptomatology, and mortality.

We unknowingly "screened" all NICU infants for elevated levels of serum benzyl alcohol (Bz-OH) over a three-month period. The fortuitous "screening" procedure resulted from the interference by Bz-OH with a routine blood CO2 assay used for all infants; validity was proved by (1) replication of the interference pattern with Bz-OH or benzoic acid; (2) confirmation of elevated benzoic acid levels in serum in two of the four screening positive infants tested but not in control infants, and (3) disappearance of the interference patterns when Bz-OH solutions were discontinued in affected infants. Screened Bz-OH-positive infants were compared to screened negative control infants, matched for weight (less than 1000 g), severity of RDS (on respirators), exposure to Bz-OH, and survival for longer than 48 hours. Intraventricular hemorrhage (IVH), metabolic acidosis appearing prior to IVH, and mortality were increased in Bz-OH-positive infants (P less than .05 in each case); hyperbilirubinemia and thrombocytopenia were not. Gasping respirations were not a major symptom. It is concluded that Bz-OH poisoning was a major cause of morbidity and mortality in NICU infants weighing less than 1000 g at birth during the three-month screening period. A retrospective review of patient records covering a 16-month period showed significant improvement in the survival rate of infants weighing less than 1000 g following the discontinuation of Bz-OH solutions.