Delayed delivery of intravenous gentamicin in neonates: impact of infusion variables

To investigate infusion variables that delay delivery of gentamicin through neonatal infusion lines.

Infusions were set up to simulate administration of gentamicin to neonates. The primary infusion was 10% dextrose (Baxter Colleague pump). A syringe driver was used to deliver a coloured marker via the T-connection over 35 min followed by a 1 ml normal saline flush over 35 min. Effects of dextrose concentration, primary infusion rate, dose volume and angle of the primary line were investigated. Gentamicin adsorption to in-line filters (Poisdyne Neo) and administration protocols from different neonatal intensive care units were also investigated.

Low dose volumes (<0.4 ml) infused into slow-flowing glucose (dextrose) lines (3.8-4 ml/h) did not mix well at the T-connection. Coloured solutions formed an upper layer that moved in a retrograde direction towards the primary infusion bag. Gentamicin did not adsorb onto Posidyne Neo filters. Comparison of infusion protocols for gentamicin administration showed that slow infusion (30 min) into slow-flowing lines (4 ml/h) containing 10% glucose gave low recovery of drug during the infusion (<30% of intended dose).

Poor mixing at the T-connection appears to be the cause of delayed and/or incomplete gentamicin delivery for low dose volumes and slow infusion rates.