Automated dispensing cabinets (ADCs) are perhaps the most ubiquitous technology in US hospitals. According to Pharmacy Purchasing & Products’ 2015 State of Automation survey, 92% of US hospitals utilize ADCs.1 In 2013, Allina Health’s ADC vendor informed health system leadership that the version of the cabinets in use would be sunset, meaning that the institution would be required to upgrade all of the existing cabinets with new hardware or consider changing vendors for this essential technology—a daunting prospect requiring extensive effort either way.
Careful consideration of the costs and workflow implications of these two possibilities was paramount to making an informed choice. After reviewing the options, the health system chose to acquire all new ADC technology. The implementation of more than 500 new cabinets in highly diversified hospital, clinic, and off-site environments largely owes its success to an emphasis on active communication among all stakeholders, appropriate staff training, and gaining staff buy-in.
Allina Health is one of the largest health systems in Minnesota and western Wisconsin. It comprises 13 hospitals, the largest of which is Abbott Northwestern, located in Minneapolis. ADC use varies widely within Allina Health due to the number of facilities, including several critical access hospitals and other secondary and tertiary care centers of varying sizes.
Abbott Northwestern has 640 licensed beds and a staff of over 3000 health care professionals that provides care to an average of 500 patients per day. The hospital had a long-standing history of more than 20 years using its previous ADCs. When it was time to begin the change process in early 2014, the first step was to establish an executive steering committee, which sought feedback from all of the relevant hospital disciplines to identify desired ADC features.
Key ADC Features
Several factors influenced the ADC choice. The requirements and workflow considerations of nursing and pharmacy are the most important factors that should drive purchasing decisions. For nursing, the ability to interact with an ADC without physically standing in front of it was a key selling point of the ADC we chose. Nurses can queue medications from any computer, including from the patient bedside, which virtually eliminates long lines at the cabinets. In addition to reducing the time nurses spend on medication passes, this feature makes documenting waste and returns more efficient as well. For pharmacy, key selling points included the reporting capabilities and the safety benefit of using NDC bar coding for item confirmation during cabinet restocking. Also, changing cabinets allowed the facility to significantly increase capacity without increasing footprint.
Based on nursing, pharmacy, and other end users’ feedback, along with the leadership team’s previous experience at other health systems, the team was able to make informed decisions regarding the various options for cabinet setup. For example, the team decided to employ only locked, lidded drawers to secure medications, which provide a higher level of security, and ruled out some of the options offered as alternatives to routine drawer dispenses.
Transitioning to a Centralized Server
The Allina executive steering committee decided that the process of changing ADCs provided the ideal impetus to examine the institution’s server setup. The decision was made to transition from individual servers at each site to installing a centralized server for all ADCs that would be maintained by the corporate office, thus removing some of the IT burden from the smaller sites. Responsibilities, such as user setup, troubleshooting, and database maintenance, could then be handled by corporate IS security or other IS personnel.
One of the early planning challenges was obtaining consensus on a standard set of configurations from all sites that use this technology. With different cultures and workflows for accomplishing tasks, merging into one platform initially seemed nearly impossible. The method that proved most successful in this endeavor was convening a best practice session at Allina’s corporate headquarters. This meeting brought together representatives and content experts from each site, as well as representatives from our new vendor. The vendor’s participation was valuable, as they were able to answer questions about the new technology and provide configuration suggestions and solutions not previously considered. Prime topics of discussion during the session included user templates, controlled substance handling, time-out settings, and central pharmacy workflow. The meeting provided a greater understanding of how all stakeholders were using the current ADCs, so that we could move forward with configuring new ADC technology that would accommodate all needs.
Somewhat ironically, a major lesson learned from the meeting was that not all configurations needed to be uniform across the health system. Coming together showed us that while there were elements we could and should standardize, other elements should be left to individual sites to maintain. The meeting demonstrated that while vendor representatives are helpful in identifying important issues, ultimately, Allina staff members must be responsible for developing the solutions, as they are the most familiar with the unique workflows of each site.
Following the large group best practices session, representatives returned to their respective sites to develop deployment plans for their hospitals. Weekly meetings with the new vendor, as well as with pharmacy and nursing leadership, facilitated communication of important information and the receipt of feedback.
The first phase of deployment for Allina Health, which began during the summer of 2014 and continued through November of 2015, included replacing the existing 319 ADCs system-wide. Abbott Northwestern Hospital alone had 85 ADC locations that were addressed as part of Phase I. Each inpatient area has at least one ADC, and most have between two and five. In addition, the institution has several procedural areas, clinics, outpatient procedural areas, and emergency departments, all of which require ADCs or anesthesia workstations. A unique factor for Abbott is the offsite emergency room located 19 miles from our main campus, but supported by the main campus. Addressing the needs of each site was essential and time-consuming. Having such a wide variety of dispensing locations forced the team to thoroughly understand and explore the options available from the ADC vendor. Involving nursing leadership, respiratory therapy leadership, anesthesia leadership, anesthesia support personnel, and pharmacy leadership early on in the design and build process was essential to the successful implementation.
As Abbott began replacing ADCs, key issues and lessons learned included:
Location, access, and medication security. These aspects are tightly entwined and had to be considered together. In addition to planning where each new cabinet would be located, the team needed to determine who required access to each ADC. Initially, the team assigned nurses to the ADCs in the patient care unit where they primarily work, but it quickly became apparent that this was too restrictive. Like many institutions, nurses at Abbott float between inpatient areas within similar communities, so the team ultimately had to create megacommunities to provide the nursing staff with access to ADCs in similar nursing units. For example, medical/surgical inpatient units were combined into one large group, and critical care service was combined into another. Some nurses, such as those in the float pool and rapid response team, frequent all inpatient units, so they required broader access. Initially, the team questioned whether this less restrictive access would be accompanied by decreased security, but the changes actually allowed for more complete and robust reporting and the ability to more accurately track suspicious activities.
At the beginning of this process, Abbott employed 25 distinct user templates for a variety of users. One of the goals established in the best practices session was for each hospital to pare down their number of user templates to only those that are essential. Some templates were abandoned as they were no longer needed; others were forfeited for the sake of system standardization. Ultimately, we reduced our user templates to 20 that are shared across the system.
Staff training. Training also presented challenges, which were ironed out over time. The new vendor provided online learning modules that were coupled with live, hands-on, in-class teachings by vendor trainers. Initially, the vendor recommended a train-the-trainer approach, whereby superusers from each community would be identified and serve as trainers for their respective units. The leadership team decided that this was not feasible and requested that the vendor perform all of the training. The team brought in trainers who led around-the-clock classes for all shifts of end users. (Ultimately, we did employ superusers during the go-live period as sources of assistance, but they did not serve as trainers.) By having a dedicated group of trainers, the team believed the vendor could deliver consistent messages regarding the functionality of the system. However, the longer-term implications of this decision, and the value of Allina staff who were trained by the vendor to be trainers, became apparent when new staff was hired, and no trainers or superusers were in place to teach the new hires how to use the equipment.
Another issue that was illuminated during the first round of training was the trainers’ lack of familiarity with the hospital’s operations and policies and procedures (P&Ps). Numerous questions were asked during training sessions that could not be answered in a timely fashion. This was corrected during training for the second phase of installations; Abbott pharmacy staff members were available to answer questions pertaining to Abbott-specific P&Ps.
We also learned that in addition to formal training sessions, providing offline cabinets (without active medications) in centralized locations is highly beneficial. This allows pharmacy, nursing, and anesthesia personnel to avail themselves of practice with the ADCs and anesthesia workstations during their downtime.
Problems for Pharmacy
For pharmacy, implementation was protracted; the department was operating both the old and new ADC systems for 4 months during the conversion process. This led to several issues. For example, monitoring controlled substances became difficult because the new cabinets could not communicate with the old ones. Monitoring and reconciling transactions became much more time-consuming, as two different systems needed to be reviewed. Refilling the ADCs also was difficult; the two systems utilize different processes, making it impossible to mirror the pick process in both ADC environments.
The time commitment for pharmacist checking also increased. The previous ADC system employed a wireless, barcode-based pick/check system, which allowed a pharmacist to check one item at a time in an expeditious fashion and electronically document their check. This allowed a single pharmacist to check an entire ADC restock batch. The new ADCs relied on the end user to sort through medications on the pick list in order to complete the validation; this process required more pharmacists to check the ADC batch restocks.
The team addressed these issues by adjusting pharmacists’ hours and asking decentralized pharmacists to help check ADC replenishment batches. In addition, pharmacy leadership adjusted pick schedules, and pharmacy transitioned to a once-daily refill model by adjusting par levels and schedules. These changes allowed us to moderate our batch size so that we could have one or two pharmacists check a batch.
Phase II of implementation, which began in June of 2015, remains ongoing and should be complete by the end of March, 2016. Phase II involves the installation of about 150 new procedure area ADCs health system wide. This includes the addition of 65 anesthesia workstations to Abbott Northwestern.
Two overarching themes have permeated our efforts and helped make our implementation successful:
Intraprofessional communication is essential. Understanding the requirements and expectations of each end-user group is paramount to optimizing the final configuration. Bringing end users, pharmacy, anesthesia leaders, anesthesia support personnel, and nursing leaders on board early in the process facilitates communication between all parties, which helps avoid pitfalls during implementation. Such communication does not end when implementation ends. The early intraprofessional communication led to the development of relationships that continue to facilitate discussions and enact workflow and system changes as issues arise. Abbott Northwestern learned during the first phase of go-lives (we have had 10 so far) that superusers are invaluable, not only for educating and supporting colleagues on new technology use, but also for providing a sympathetic ear to coworkers who become frustrated by change.
Embrace struggle. Staff members had to become comfortable with the fact that not everything was going to go smoothly and that struggles were inevitable. The staff was comfortable with the previous ADCs; they had a long history with them and had acquired expertise in using the cabinets. Implementing a new system required starting from square one. Helping staff set realistic expectations about the process was critical to success.
In addition, it was important for team leaders to determine if staff complaints were related to resistance to change or to true deficits in the new ADC system, so they could be addressed appropriately.
Switching to new ADC technology may seem a daunting task, particularly for a large health system. Nonetheless, early and ongoing communication between leadership and end users helps to ensure a successful implementation, as does an emphasis on training. Perhaps most important is helping staff set realistic expectations. While it is reasonable to strive for perfection in the end product, the scope and process are too complicated to proceed perfectly at every step along the way. Encouraging staff to embrace and become comfortable with the discomfort associated with change helps diminish frustration. In the end, multiple disciplines working together—including nursing, anesthesia, respiratory, and pharmacy—allowed us to address issues that arose quickly and safely, ensuring that switching ADC systems ultimately delivers improved patient care.
Matthew Ditmore, PharmD, BCPS, is the lead pharmacist of central pharmacy operations and automated dispensing technology at Abbott Northwestern Hospital, Minneapolis. He received his PharmD from the Massachusetts College of Pharmacy and Health Sciences in Boston and completed a 1-year pharmacy practice residency at Abbott Northwestern Hospital. Following residency, Matt was a staff pharmacist at St. Joseph’s Hospital in St. Paul working with the emergency medicine and critical care teams. He returned to Abbott in January, 2013, as a staff pharmacist working on the emergency medicine and critical care pharmacy teams until he transitioned to the lead pharmacist role in early 2014.
Enter our Sweepstakes now for your chance to win the following prizes:
Just answer the following quick question for your chance to win:
Entries are limited to one entry per person in any active sweepstakes.
Thank you for your entry.