Background The methylxanthine caffeine reduces the frequency of apnoea in prematurity and the need for mechanical ventilation. The pharmacokinetics of caffeine in preterm infants shows inter- and intraindividual variability. Therefore therapeutic drug monitoring (TDM) of caffeine is used to optimise individual dosing to prevent toxicity and treatment failure. Because blood sampling in preterm neonates is an invasive procedure, the question arises whether routine drug monitoring is necessary.

Purpose To determine the value of TDM of caffeine in preterm neonates.

Materials and methods A retrospective study was conducted at Sint Franciscus Gasthuis, Rotterdam, The Netherlands. Preterm neonates treated with caffeine were identified in the period January 2008 to June 2010. Patients with at least one plasma caffeine level determination were included. The medical charts of preterm neonates with a caffeine level > 30 mg/ml were screened for adverse events.

Results A total of 601 caffeine plasma levels were measured in 149 patients. The average dose was an induction dose of 20 mg/kg caffeine citrate and a maintenance dosage of 10 mg/kg/day. Plasma caffeine levels were between 10 and 25 mg/l in 86.5%, < 5 mg/l in 6% and > 25 mg/l in 6.8%. 1.8% of the plasma levels were >30 mg/l (range 30.2–37.4 mg/ml). In one patient in the subgroup of 11 patients with a plasma level >30 mg/l tachycardia was recorded in the medical chart as an adverse event.

Conclusions The majority of preterm neonates attain plasma levels between 5 and 25 mg/l if a standard dose is used. In the subgroup of patients with levels > 30 mg/l only one adverse event was recorded. Based on these results routine TDM is not necessary in preterm neonates.