Intravenous (IV) dose preparation has traditionally been a human-driven process, with a pharmacy technician preparing the dose and a pharmacist verifying its accuracy. However, manual preparation is rife with opportunities for error. The Institute for Safe Medication Practices states that utilizing the syringe-pullback method for verification is less than ideal, as it relies on the preparer’s memory of how much ingredient was drawn up. Automated IV workflow systems have the capacity to decrease error potential.1 In addition to improving safety, several workflow, tracking, and resource benefits can be realized when adopting an automated IV workflow system.
The Medical University of South Carolina (MUSC) is 700-bed academic medical center served by 17 pharmacy locations, including inpatient, outpatient, chemotherapy, nuclear, and mail order services. Prior to 2011, all our cleanrooms operated under a manual process for IV compounding. Concerns about IV safety, workflow, and cost drove us to investigate better options.
Ensuring a Robust System
Our goal in assessing automated IV workflow options was to identify the best system to close the gaps in the manual IV process. The solution had to be flexible enough to universally improve our processes across different practice settings. As this type of technology is relatively new, we were wary of implementing a solution that was not yet ready for real-world use. We were particularly concerned that a new system may introduce additional and perhaps unforeseen challenges into the equation.
To examine the potential failure points in each method of IV preparation, we conducted a failure mode effects analysis of both the manual and automated workflow styles. Because of the number of pharmacies we operate, it was important to analyze data in several different contexts to ensure all areas of potential error were identified. Our analysis revealed a number of potential failure points in the manual process that could be mitigated via an automated IV workflow system (see Table 1).
In evaluating the various market options, our primary consideration was identifying a system that offered a high degree of scalability. With eight IV compounding areas on the MUSC campus, we envisioned a phased implementation approach that when complete would be linked on a single server. Additionally, we needed something that would support the significant customization required to accommodate adult and children’s hospitals, inpatient and outpatient settings, and both hazardous and non-hazardous workloads.
From the start, we intended to utilize the automated IV workflow system for all IV dose types, not limiting its use to just chemotherapy or high-risk orders. The goal of including all IVs was aggressive, but necessary, as any dose can become a high-risk preparation if the wrong ingredients are used. After evaluating the capabilities of available automated IV workflow systems, the solution that best met our requirements was identified.
Gaining Administration Support
When approaching management to seek support and funding for an automated IV workflow system, we focused on two benefits: increased medication safety and reduced waste. A significant portion of our IV workload involves high-cost or short-supply medications, especially chemotherapy. Due to recent widespread shortages, every vial is precious. Ensuring orders are prepared correctly each time—thus reducing costly waste from incorrectly prepared drugs—was a high priority.
Our analysis did not suggest additional FTEs would be necessary to operate the system. Rather than having a pharmacist stationed in the IV room to verify medications, the technology would allow the pharmacist to verify doses from any computer, utilizing information and images captured during the dose preparation process; even the hazardous preparation IV rooms could be staffed with pharmacy technicians only. For locations operating under a single pharmacist, we could leverage a second pharmacist to double-check the preparation remotely, in real time.
Because the system stores preparation records electronically, information and images could be retrieved long after the dose is administered. The ability to query this data on a large scale offered previously impossible data analysis opportunities. We were fortunate to have the opportunity to compare live demonstrations of different automated IV workflow systems at various hospitals. After visiting sites and experiencing the real-world application of the available systems, the choice, and the utility of such a system, became clear.
When adopting new technology, staff must clearly understand the reasons for change to ensure their acceptance. For example, an experienced technician may question the necessity of an automated IV workflow system. Assure the pharmacy team that the need for technology is not a reflection on their performance; rather, it is a critical safety initiative to improve an error-prone manual process. At MUSC, our experience has been that after demonstrating success using the technology, and witnessing those benefits firsthand, staff members quickly appreciate the additional tools that allow them to do their job more effectively.
Training is the cornerstone of any process implementation. In-depth education prior to and throughout the implementation period ensures that the system is operated as intended and discourages the staff from developing workarounds. We engaged all levels of staff early in the process, and their feedback was crucial to appropriate system design. End users assisted in testing as we brought the system on-line, and were motivated to provide constructive criticism. As superusers spent more time testing, their confidence in the system increased and spurred buy-in from other staff.
Prior to the go-live, we set up demonstration workstations in each pharmacy to provide hands-on training. Staff was required to process a certain number of test doses to demonstrate baseline proficiency. Starting one week prior to go-live, we switched to a parallel process—making doses using both the automated system and the manual syringe-pullback method. This approach allowed both the preparing technician and the verifying pharmacist to simultaneously compare the possible failure points between methods. Staff continued to use both systems concurrently throughout the first week of go-live. Workload indicators were disseminated daily to staff during the implementation phase, highlighting system capabilities, workflow efficiencies, and general user performance to encourage compliance over the long-term.
Dual Benefits of Bar Coding in the IV Room
Utilizing bar code verification in the IV room reduces the error potential associated with manual processes through enhanced product verification, while also providing dose-tracking capabilities. Accurately tracking a product’s movement through the drug distribution system is nearly impossible without automation.
Bar coding provides a vector for both safety and security throughout the medication distribution process. Not only are pharmacists and technicians reassured that they have selected the right ingredients prior to preparing a dose, the software also can track delivery to the destination. Since no additional software or labeling is required, tracking seamlessly integrates into the workflow.
Advances Using IV Workflow Automation
No system can cover every single dose, but we believed a goal rate in the high 90% range achievable. By the end of the first week in our initial implementation site (MUSC Children’s hospital pharmacy), 85% of IV doses were produced and verified via the automated IV workflow system. Today across our seven live locations, we prepare over 98% of all IV doses using this technology. Items produced outside of the technology include products without manufacturer bar codes, manufacturer bar codes that fail to scan due to functional constraints (eg, curvature of the vial), investigational drugs, and medications with variable concentrations per package, such as Factor IX.
Adopting an automated IV workflow system in our cleanrooms has significantly improved our error rate. Using the manual process, about 4% of doses presented for verification were rejected for correction. After implementation of the technology, our rejection rate has dropped to an average of 0.7%. Both pharmacists and technicians operate with a higher level of confidence, and enjoy the ability to reference dose preparation information even after it leaves the pharmacy.
Prior to automating our hazardous IV workflow, the process for pharmacists to check doses was error-prone and inefficient. The technician preparing the dose had to pause before the final preparation phase, waiting until the pharmacist could verify the amount drawn up to avoid pullback method error. The pharmacist had to garb up to enter the cleanroom, and both the pharmacist and technician had to wait on the other to complete their respective tasks.
The automated IV workflow system captures images of the actual products used to prepare a dose as they are used, allowing the pharmacist to verify the medication contents with certainty without entering the cleanroom. The pharmacist and the technician can carry out their duties concurrently, with no need to wait on the other before proceeding to the next step. While the pharmacist is verifying a dose, the technician can move on preparing the next medication (see online-only Figure at www.pppmag.com/ivwm).
This technology also permits redistribution of staff resources, further increasing workflow efficiencies. For example, if a pharmacist is called away on a code and an emergent medication must be verified in the pharmacy, a pharmacist in another facility in the health system can perform this task remotely via Web interface, ensuring the patient receives the drug in a timely manner. This is particularly helpful in rural hospitals without 24-hour capabilities, as a pharmacist can verify doses remotely during off hours (provided state regulations allow).
Automated IV workflow systems provide remarkable data-capturing opportunities. The capacity is so robust that our initial data reviews resulted in an entirely new set of questions surrounding our workflow. For example, there are meaningful differences between doses sent in a carrier bag versus those dispensed in a syringe, but we could never elucidate them. The software makes delineating differences among dose characteristics simple, so one can easily aggregate data too time-consuming to compile manually. The technology also tracks which staff member performs which tasks, where the work is done, and when. This data facilitates more efficient evaluation of staff utilization and assists in efforts to improve staff productivity.
Given the number of pharmacies in the MUSC system, the value of standardizing dose preparation was clear. Preparation points, such as what diluent to use and how to handle overfill, were the first targets. Using an automated IV workflow system to ensure staff adherence helped us achieve process consistency.
Standardization began with identifying best practices. Admittedly, adult, pediatric, and hazardous dose preparation steps are different, but upon close examination of the preparation steps, many similarities become apparent. Workflow around preparation images presented the largest variable. Two questions needed to be answered: At what point(s) in the process should images be captured? and What information should these images depict? In addition to recommendations from the vendor, we worked with our pharmacists to ensure images captured the most pertinent information and that photographing that data was feasible in the technicians’ workflow. We also specified which direction syringes should face in the photos. Standardizing operations to that level delivered the consistency our staff needed to be comfortable preparing doses in any hood.
The ability to track doses across our enterprise is remarkably beneficial. Being able to locate a medication at any point in the preparation or distribution process has created a new hospital culture focused on decreasing waste. Pharmacy delivery time has changed from a subjective perception to objective data, and pharmacy and nursing staff work together to quickly locate missing doses. As a result, our monthly average missing IV dose requests in the children’s hospital has been cut by 50%.
Further savings have been realized as a result of the improved accuracy of doses prepared. Now that verification begins before a vial is even opened, scanning each component ensures each medication is correct and accurately prepared, thus reducing wastage from improperly prepared doses that must be discarded.
Patient safety is the most significant benefit of an automated IV workflow system, and understanding and controlling variables throughout the preparation process creates consistent outcomes. Fewer wasted or lost doses offer additional costs savings to the organization. The versatility of our automated IV workflow system allows for a smooth workflow, scalable to meet the needs of a small community hospital or a large medical center.
However, there is no panacea in medication safety. Bringing technology to a pharmacy with a poor workflow will only highlight existing problems, and may even exacerbate them. Fundamentals—such as proper technique, appropriate storage, and staff training—are prerequisites to technology implementation. Integrating IV technology into an already successful workflow can enhance safety, control workflow, improve tracking capacity, and permit more effective resource distribution.
Jeff Brittain, PharmD, BCPS, a clinical pharmacist in medication policy and informatics at the Medical University of South Carolina in Charleston, manages MUSC’s automated IV workflow system and also serves as a South Carolina College of Pharmacy preceptor. He received his PharmD from Wilkes University in 2005 and Board Certified Pharmacotherapy Specialist credential in 2009.