In pediatric units, most of the intravenous medications are prepared by the attending nurse at the bedside that can be affected by an error margin, so can be imprecise. Despite the possible consequences of imprecise medications administration, published studies on the topic are scarce. The main objective of this study was to measure the difference between the prescribed vancomycine concentration and the actual concentration measured in the medication administered to the patient. The secondary objective was to determine which step in the preparation was linked to the difference in concentrations. It was a prospective study, setting in a pediatric and neonatal university hospital intensive care unit. Over a 3-month period, an aliquot from every preparation for continuous infusion of vancomycin, made at the bedside by a nurse, was collected and the modalities of the preparation noted. Vancomycin concentration was measured by high performance liquid chromatography. Sixty-four preparations, accounting for 24 patients (gestationnal age: 67 ± 75 weeks, weigh: 4.8 ± 6.5 kg) were included. Vancomycin concentrations ranged from 3.33 to 60.0mg/mL. Measured concentration were in mean 7% smaller than prescribed concentration (P<10(-3)), with a large confidence interval (75.8%-120.4% of the prescribed concentration). Imprecision the preparations was much higher than this admitted for manufactured preparation. We could not highlight any factor related to the difference in concentrations, but one third of the preparation did not respect all the ISO 7886 standards for syringes use. Bedside vancomycin preparations, like preparations for other molecules, are far more imprecise than industrial intravenous medications. Our results urge that all pediatric intravenous medications should be made only by manufacturers or pharmacists. However, it also urged clinical studies, in parallel to pharmacodynamic and pharmacokinetic studies, to make intravenous treatments as accurate as they should be.
2011 Elsevier Masson SAS. All rights reserved.