Electronic versus handwritten medication prescriptions improves safety in neonatal intensive care units

Background Neonates receiving complex dosing regimens are at an increased risk of iatrogenic events. Electronic prescribing (EP) has been proposed to reduce medication errors compared with handwritten prescriptions (HP). This study aimed to determine in a specific setting whether EP with guiding clinical support was less error prone than HP with access to literature. Methods A prospective crossover study involved 77 neonatologists in eight French neonatal intensive care units (NICUs) and two Spanish NICUs. Each neonatologist wrote two prescriptions (four drugs each) by hand and electronically. The order was randomly assigned for 63 prescribers. In two of the NICUs, no computerized system was used, so 14 prescribers wrote only HP. Prescribers were timed and allowed to consult any documentation. Prescriptions were qualitatively analyzed against established reference standards. Each prescription was required to include the dosage form, unit dose, number of administrations, and specific preparation/administration instructions, as would be done in actual practice. Results The median time [IQR] to write a prescription was 8.0 minutes for HP [4.9-11.7] and 3.5 minutes for EP [2.9-4.3] (p < 0.0001). Overt errors were significantly more common with HP versus EP (145/604 prescriptions versus 3/468; p < 0.0001) and 382 prescriptions were found to be unclear writing (none in EP). All types of errors were observed in HP, including calculation errors such as incorrect cross multiplication, incorrect units, conversion errors, and errors in dosage or interval. Conclusion The use of a computerized system specifically designed for neonatology significantly reduced medication prescription errors. It also significantly reduced prescription time.