Impact of Smart Pump Interoperability: Reducing Alerts and Improving Safety

Purpose: Interoperability between smart pumps and the electronic medical record (EMR) closes the loop on medication safety, helping ensure the right patient, right medication, right dose, right route, right time, and right documentation. Infusion orders are auto-programmed into the pump, reducing the need for manual programming and associated risk of error, and infusion rates and volumes are auto-documented into the EMR. This clinical decision support system includes alert triggers for programming outside defined dosing limits, tracking software to identify practice trends, and visibility to infusions in real-time to monitor and improve patient safety. While alert limit triggers help safeguard infusion delivery, if they are not driven by hospital evidence, they can contribute to alert fatigue and drug library non-compliance. This project evaluated the impact of smart pump interoperability on clinical practice by measuring and evaluating drug library alerts, overrides and compliance rates at an acute care hospital.
Description:In Sept 2018, wireless infusion pumps were implemented with dose tracking software in an acute care hospital. Data were collected and analyzed on drug library compliance, alerts, and overrides using infusion management software. There were 3 data collection periods: baseline (October - December 2018), post-EMR integration (October - December 2019), and post-drug library intervention (APril - June 2020). Infusion data was sent via secure file transfer protocol to a secure central server protected by firewalls, HIPAA de-identified, under software licensing and data use agreement.
Baseline data showed drug library compliance of 96%, with an alert override response rate (alerts that resulted in an override) of 12% and override frequency (total overrides/total drug library deliveries) of 1.45%. EMR interoperability went live in August 2019 and included changes to the pump-based drug library limits to match the EMR formulary. Post-EMR integration, drug library compliance increased to 98%. However, the alert override response rate increased to 58% with an override frequency of 4.19%. Eighty percent of the overrides were associated with norepinephrine and phenylephrine due to the addition of soft minimum dosing limits to match the EMR. However, these infusions are commonly weaned off below soft limits, making these alerts excessive and clinically non-credible. Subsequently, the drug library was modified to eliminate these soft minimum limits, virtually eliminating overrides associated with these infusions. As a result, alert override response rate decreased to 9%, override frequency further decreased to 0.33%, and drug library compliance sustained at 97%.
Outcome: Smart pump interoperability was successful, achieving 91% auto-programming compliance in the ED and 77% hospital wide immediately after go-live. Evaluating data at routine intervals identified opportunities to optimize the drug library; reducing overrides 77% and reducing alert override response rate to a remarkably low 9%, compared to rates of 74%-76% reported by Marwitz et al. Minimizing overrides also helped sustain high levels of drug library compliance of 96-98%, versus 81% national average. Smart pump interoperability impacted patient safety; one year post integration (as of 10/28/20), there were no reported errors related to smart pumps. A limitation is that our outcomes cannot be extrapolated to other hospitals.