Analysis of Event Logs from Syringe Pumps

Abstract

Medication errors occur in approximately one out of five doses in a typical hospital setting. Patients in the intensive care unit (ICU) are particularly susceptible to errors during the application of intravenous drugs as they receive numerous potent drugs applied by syringe pumps. The aim of this study was to analyse the effects on potential harmful medication errors and to address factors that have potential for improving medication safety after the introduction of a standardized drug library into syringe pumps with integrated decision support systems. A team of physicians and nurses developed a dataset that defined standardized drug concentrations, application rates and alert limits to prevent accidental overdosing of intravenous medications. This dataset was implemented in 100 syringe pumps with the ability to log programming errors, alerts, reprogramming events and overrides ('smart pumps'). In this retrospective pilot study, all pump-related transaction data were obtained from the pump logs, by downloading the data from the pumps, covering 20 months of use between 1 April 2008 and 30 November 2009. Patient data were gathered from the electronic patient charts. The study was performed in a cardiothoracic ICU of the Charité University Hospital, Berlin, Germany. A total of 7884 patient treatment days and 133,601 infusion starts were evaluated. The drug library with the features of the dose rate was used in 92.8% of the syringe pump starts, in 1.5% of the starts a manual dosing mode without the use of the drug library was used and in 5.7% of the starts the mode 'mL/h', without any calculation features, was used. The most frequently used drugs were vasoactive drugs, followed by sedation medication. The user was alerted for a potentially harmful overdosing in 717 cases and in 66 cases the pumps were reprogrammed after the alert. During the early morning hours a higher rate of alarms was generated by the pumps, compared with the rest of the day. Syringe pumps with integrated safety features have the capacity to intercept medication errors. The structured evaluation of the bedside programming history in log recordings is an important benefit of smart pumps, as this enables the users to obtain an objective measurement of infusion practice, which can be used to provide team feedback and to optimize the programming of the pumps. Further research will show if the combination of these data with physiological data from ICU patients can improve the safety of pump-driven intravenous medication.