Optimizing the Use of Dose Error Reduction Software on Intravenous Infusion Pumps Free

Our institution is a 77-bed specialty hospital that provides care for pediatric patients with oncologic, hematologic, and other catastrophic diseases. Our institution sees 8600 patients annually in inpatient and outpatient settings with an average of 7200 infusions weekly. Most patients admitted to the hospital receive IV infusions during their admission, and 52.6% of admissions include administration of IV chemotherapy during that admission. Guardrails on Alaris pumps (BD, Franklin Lakes, NJ) is the infusion pump DERS that is used at our institution. These pumps had interoperability with the EHR (Oracle Health EHR, Oracle Corporation, Redwood Shores, CA; formerly, Millennium, Cerner Corporation) at the beginning of this project and transitioned to interoperability to another EHR (Epic, Verona, WI) in October 2022.

At our institution, DERS was implemented in 2002 for select drugs. Interoperability between the EHR and smart infusion pump was implemented in 2017, and it decreased the number of steps for the end user to program a medication on an infusion pump by ∼20 steps. In preparation for smart pump–EHR interoperability implementation, a focus point was to update the drugs built in DERS. At that time, use of DERS was limited to only high-alert drugs, and so only 8.9% of all infusions were being programmed using DERS. The drug library update in preparation for smart pump–EHR interoperability increased compliance from 8.9% to 40.6%. After implementation of smart pump–EHR interoperability, DERS compliance increased to ∼50%. A review of DERS compliance in 2019 revealed that compliance remained at 46% despite 2 years of active smart pump–EHR interoperability. This low compliance indicated a high rate of end users bypassing the smart pump safety features and ultimately led to the formation of an interdisciplinary workgroup to optimize the benefits of this technology.

An interdisciplinary workgroup was formed, comprising members from nursing, nursing education, medication safety, informatics, and clinical research pharmacists. The group was tasked with deciding the aims, key drivers, and appropriate interventions for the project. Three key drivers were identified, including informed staff, engaged staff, and an accurate smart pump library (Fig 1). The Model for Improvement framework was used to guide this improvement project, and iterative Plan-Do-Study-Act (PDSA) cycles were used to plan for interventions during this project.^19,20  PDSA cycle descriptions and lessons learned from each cycle are outlined in Table 1.

The first PDSA cycle focused on evaluating smart infusion pump data to identify medications that were firing the most pump alerts to the clinician. The top 20 medications presenting alerts were evaluated to identify potential drug library changes (Supplemental Fig 4). After review by the workgroup, the drug library settings of 10 medications were changed in April 2019 (Supplemental Table 2).

Because of competing priorities, the interdisciplinary workgroup did not meet again for ∼18 months. Once reconvened, the workgroup began to plan the next action steps. PDSA cycles 2 to 4 focused on education, including the importance of using DERS versus basic infusion mode, instructions for using DERS for blood products and flushes, and training for when smart pump–EHR interoperability fails. PDSA cycle 5 included implementation of DERS monitoring and education during monthly walking safety rounds, and PDSA cycle 6 incorporated reporting of unit-specific compliance data to nursing leadership. PDSA cycle 7 integrated implementation of quarterly drug library updates; updates to the drug library were based on alert override data, feedback from nursing, event reports, and medications being built into the future EHR that were not already present in the pump library. Finally, PDSA cycle 8 reinforced previous nursing education through the creation and completion of an interactive nursing competency.

Weekly DERS compliance rates were monitored from July 2018 to December 2022. DERS compliance was calculated by reviewing usage reports of medications and IV fluids from the vendor’s tool (Integrated Analytics Knowledge Portal by BD). Blood products and IV flush entries were excluded. The compliance rate was defined as the number of DERS infusions per total infusions administered. The balancing measure was defined as pump alerts per 100 infusions and monitored from July 2018 to December 2022. The weekly frequency of DERS pump alerts was determined by calculating the number of alerts per 100 DERS infusions based on vendor-supplied data. This metric was selected as the balancing measure to ensure that alerts did not increase because of efforts to improve compliance, particularly as more drugs were added to the drug library and more nurses began to use DERS.

Weekly compliance rate percentages over time were analyzed by using a statistical process control (SPC) chart (Laney p’chart) in Excel. Weekly alerts per 100 DERS infusions over time were analyzed using an SPC chart (u-chart) that was created with SPC for Excel. Accepted statistical conventions for SPC were followed.²¹  The average (center line) was recalculated when a special cause shift was detected for 8 data points above or below the centerline.²² 

This project was reviewed by the institution’s institutional review board and found to be a quality improvement project.

After the initial DERS library update took place in April 2019, our outcome measure of DERS compliance increased from 46% to 61% (Fig 2). PDSA cycles 2 to 4 showed only incremental improvements. PDSA cycles 5 and 6 were implemented in proximity, and an increase in compliance was noted. The average DERS compliance was recalculated at 67%, identifying PDSA cycles 5 and 6 as the reasons for the system change. PDSA cycle 7 and subsequent drug library updates resulted in an increase in compliance to 73%. Compliance rates experienced a centerline shift beginning June 5, 2022, bringing the new DERS compliance rate to 78%. No increases in compliance were seen after PDSA cycle 8. The institution transitioned to a new EHR in October 2022, and DERS compliance remained consistent throughout this time.

For the balancing measure of pump alerts per 100 infusions, the baseline rate of 15.9 alerts per 100 infusions decreased to 6.4 after the first PDSA cycle (Fig 3). Pump alerts per 100 infusions continued to decrease throughout the project, with the largest decreases after PDSA cycles 5 and 6 (5.4 alerts per 100 infusions) and after PDSA cycle 8 and the EHR transition (3.9 alerts per 100 infusions).

In this improvement project, DERS compliance increased, although our balancing measure of pump alerts did not increase. Interventions were facilitated through interdisciplinary collaboration, and the greatest improvements were seen after PDSA cycle 1, with changes to the drug library based on the most common alerts, PDSA cycles 5 and 6 with targeted discussion during safety rounds and unit-specific feedback on compliance, and PDSA cycle 7 with quarterly drug library updates. Because of the increase in compliance seen with these interventions, they were incorporated into the routine functions of the interdisciplinary workgroup, which was expanded to oversee all pump issues and continues to meet routinely.

To our knowledge, this is the first published improvement project focused on improving DERS compliance in the context of smart pump–EHR interoperability. Although it’s reasonable to assume that nurses would want to use this technology because of the reduction in pump keystrokes and the EHR auto-documentation, some studies have suggested there may be an upper limit to DERS compliance after the implementation of pump–EHR interoperability.^23,24  Indeed, after implementation of smart pump–EHR interoperability at our institution, DERS compliance increased to ∼50%, but did not improve for ∼2 years until the interventions described in this report. In short, these technologies are not “plug and play” and require significant institutional investment after implementation to realize the full safety benefits.¹⁴  Continuous monitoring and improvement efforts are needed to fully realize the safety benefits of DERS with smart pump–EHR interoperability.

Our experience illustrates the importance of a continuously updated and accurate drug library on the use of the pump’s safety features because this intervention improved compliance the most. DERS compliance improvements were seen after PDSA cycle 1, with changes to the drug library based on the most common alerts and PDSA cycle 7, with quarterly drug library updates. Abboudi and colleagues modestly increased compliance from 77% to 83% over a 2-year period, with the greatest impact coming from nurses directly communicating DERS issues to informaticists making library updates. Although their specific interventions were different than our experience, the common thread for improvement is consistently monitoring and updating the underlying drug library information. Like our experience, the group did not observe meaningful improvement with education only.²⁵ 

To sustain these improvements and reach the goal of 95% compliance suggested by the Institute for Safe Medication Practices, a next step is to instill standardization between the drug library and EHR.¹⁸  As our hospital transitioned to a new EHR vendor, a thorough review of the drug library in the EHR afforded an opportunity to carefully review the drug library in the pumps, as well. Compliance remained consistent during the transition. However, data post go-live are limited, and changes to the EHR drug build process continue to be refined. Joseph and colleagues also highlight the need for complete alignment of the drug library and EHR medication orders; otherwise, interoperability may fail, and the drug library may not be used.²⁴ 

Although our work reinforced the importance of drug library maintenance, it also emphasized the magnitude of unit-specific feedback. PDSA cycles 5 and 6 included targeted discussion on using DERS during monthly walking safety rounds and reporting of unit-specific compliance data. Similarly, others have described using weekly compliance reports and education with feedback to improve the drug library and DERS compliance.³ 

Our facility cares primarily for pediatric patients with oncologic, hematologic, and other catastrophic diseases, and therefore, our patient population frequently requires high-alert drugs (eg, chemotherapy), which may make this project less generalizable to other facilities. Further, only 13.4% of institutions have implemented smart pump–EHR interoperability.⁵  However, our interventions could be implemented at institutions with DERS only, and our experience may provide insights for improvement as other hospitals pursue smart pump–EHR interoperability. Further, the substantial ongoing effort to increase compliance with DERS may have diverted our attention from other necessary improvements in infusion safety. However, our improvements appear sustainable, and this project developed new processes and collaboration to inform a more comprehensive strategy for the safety of all infusions.

Our quality improvement efforts significantly improved DERS compliance in an institution with smart pump–EHR interoperability. The most critical of interventions were related to review of most common alerting medications with subsequent changes to the drug library, targeted discussion during safety rounds, unit-specific compliance reporting, and routine drug library updates. Although features like DERS can help ensure safe medication administration, continuous improvement efforts are needed to ensure that benefits of this technology are optimized.

We thank all members of the DERS compliance group at St Jude for their collaboration and support toward this quality improvement initiative to promote safe and efficient use of smart pump technology at St Jude Children’s Research Hospital. We thank David Aguero, PharmD, MSTL, ACHIP, and Samuel Bowker, PharmD, MHIIM, for their support in optimizing the safety features for our infusion pumps, and Ben Moore, PharmD, MS, and Murad Hasan, DrPH, for their support with data analytics; Chad Compagner, PharmD, for being vital in implementing a new drug build process that includes pump drug library additions; and Jonathan Burlison, PhD, William Mabry, PharmD, MBA, and Paul Sharek, MD, MPH, for helpful feedback on drafts of the article.