Älyinfuusiopumput lääkehoidon riskienhallinnan välineenä : annosten turvarajojen pilotointi HUSin vastasyntyneiden teho-osastolla

Medication errors due to infusion pump programming errors are common in neonatal intensive care units and often lead to overdoses of medicines. Medication errors can be prevented with dosing limits set into the smart infusion pump drug library. However, alert fatigue caused by unnecessary alarms has been identified to hinder their use. To ensure the benefit of dosing limits as a defence of intravenous medication process, the aim of this study was to define the dosing limits to the drug library for certain high alert medications, and to pilot them in the neonatal intensive care unit. The study was based on the theoretical framework of preventive medication risk management. Based on the results, the suitability of the dosing limits for the use of the unit and the patient group was assessed. This mixed method study employed register-based research methods. The research data consisted of the infusion rate related medication errors (n=21) reported to the HUS HaiPro-system between January 2018 and December 2019 in the HUS Neonatal Intensive Care Unit. The data was analysed qualitatively and quantitatively to describe the infusion rate related errors, and to identify their mechanisms and contributing factors. Based on the identified mechanisms of the errors, simulated test patient cases were developed. Dosing limits were defined by a multidisciplinary expert group for certain high alert medications, and their suitability for preventing medication errors was investigated by programming infusion pumps according to the test patient cases and analysing the pump alerts. Based on the identified mechanisms of infusion rate-related medication errors (n=21) in the HaiPro-reports, 2-, 5-, and 10-fold infusion rates as well as mixing of infusion rates between medicines were established as test patient cases. As a result of the tests (n=226), the infusion pumps did not alert when programming normal infusion rates (n=32) and 73% (n = 70/96) of the erroneous 2-, 5-, and 10-fold infusion rates were prevented. 10-fold infusion rates were inhibited in all cases (n=32). Interference of infusion rates between medicines was prevented in 24% (n=24/98) of the cases. According to this study, significant infusion rate related medication errors can be prevented in the neonatal intensive care unit with the multidisciplinarily defined dosing limits set in the infusion pump drug library. However, they do not prevent all the infusion rate related medication errors alone, and therefore additional defences are needed. In addition to the neonatal intensive care unit, the method used in this study to define and test the dosing limits may be applied in other pediatric units in the future. By using this method, the suitability of the dosing limits for the use of the unit and the patient group can be ensured before integration of the barrier and thereby promote the benefits of its use. The suitability of the dosing limits set into the infusion pump drug library should be assessed again after implementing the defence into the neonatal intensive care unit.