Implementing a BCMA system is no easy task, and once the heavy lifting of implementation is over, continuous vigilance and collaborative effort between pharmacy and nursing are then required to maintain high scan rates of both patients and medications. Furthermore, heedfulness is necessary to ensure that system workarounds do not whittle away at the safeguards provided by the established medication administration process. Thus, in order to maximize the utility of the system, bar code scanning of 100% of patients and medications should be the target.
At Beaumont Health System–Royal Oak (BHS-RO)—a 1070-bed, tertiary care, academic medical center with a level 1 trauma unit and a children’s hospital within the facility—pharmacists are responsible for both distributive and clinical services. To better support our specialty areas, we recently restructured, and now, instead of working out of multiple satellite locations, our pharmacists are concentrated largely on the units. Our medication distribution process consists of profiled ADCs that provide approximately 80% of PRN and scheduled medications. The remaining, patient-specific doses are provided via a 24-hour cart fill that utilizes carousel technology. BHS-RO also implemented an electronic health record (EHR) system in phases over a number of years. At present, the facility uses CPOE and eMAR systems in nearly all areas of the hospital. Likewise, BCMA is used for the majority of the 6.5 million doses administered annually, although currently it is not used intra-procedure or during codes.
Establishing Recognizable Bar Codes
The BCMA implementation process took approximately 1 year to plan and an additional 6 months to fully execute. From a pharmacy perspective, the first step was to make sure we either were purchasing unit-dose medications with manufacturer-supplied bar codes, which is the case for the majority of our medications, or that we could include a bar code during repackaging. We also determined early that a relatively small number of over-the-counter medications would not be included in the BCMA system. This decision applied to products for which the amount of effort and cost associated with repackaging and/or affixing a bar code far exceeded the risk associated with possible incorrect administration. An example of such a product is menthol lozenges, which are purchased in foil-wrapped cards of nine. Pharmacy worked with the EHR build team to exclude such medications from requiring a scan prior to administration.
Because no standards exist regarding the design of manufacturer bar codes, the next step was to determine whether our EHR could recognize these bar codes, as recognition by our carousel software did not assure recognition by the EHR. Rather than attempt to predetermine which medications might be dispensed directly out of manufacturers’ packaging, we elected to validate the bar code on virtually every product stocked in our carousels. This validation process, which took approximately 3 months and had to be completed before the system went live, involved scanning the bar code to determine whether it matched an existing NDC in our database. Products with codes that did not match an NDC were set aside for the EHR build team to analyze in order to develop a rule that would allow the scan to be recognized and linked to a known NDC.
Supplemental to validating bar codes on our stocked medications, we decided that as deliveries from our wholesaler were unloaded, our stock clerks would scan an individual package (eg, tablet, capsule, and vial) of each lot number of each medication received. If a medication was not recognized, it was not stocked in the carousels. Instead, the build team was contacted to make whatever changes were necessary to enable recognition by our EHR. Only then was the medication moved to the carousels. The stock clerks were trained on all aspects of the BCMA process, with emphasis on the importance of scanning every medication—even those scanned and validated previously—to make certain that all medications were recognizable by the EHR. This process continues today.
In addition to manufactured products, we needed to determine proper bar codes for our compounded doses (eg, large and small volume parenterals, patient-specific oral syringes) and some bulk medications (eg, metered-dose inhalers, creams, ointments). This process involved reformatting our existing EHR-generated labels to include three relatively small, 2D, order-specific bar codes. We also used this strategy when developing our patient wristbands. We believed this solution would reduce the number of situations in which the bar code was rendered unreadable by damage, and the method has proven effective.
Ultimately, having a readable bar code on 100% of our dispensed medications was critical to the overall success of the BCMA program. It does not take long for nurses to discover that a particular medication does not scan and to begin overriding procedure, a behavior that often persists even after the scanning issue is resolved. Occasionally, we still dispense a medication with a missing or unrecognizable bar code; this sometimes occurs when obtaining a medication from another institution during off hours or when the bar code is not validated prior to dispensing.
Flaws and Exceptions
The above-mentioned process of validating bar codes unearthed a flaw in our system. After initiating the process, we began receiving reports of a frequently used controlled substance not scanning properly. In order to exert strict security over controlled substances, these medications were exempt from being scan-checked by the stock clerks, so validation scans are not always performed. At the time, we received a large shipment of controlled substance product that had a new NDC, which was not yet entered in our EHR database. Almost 24 hours elapsed before we were alerted that the EHR was not recognizing the bar code. To make matters worse, the discovery occurred on a weekend, and the change was not made in the EHR until the following Monday. It took weeks for the scanning rates of that medication to return to previous levels. Now, the pharmacist who checks in controlled substances performs validation scans.
To address another common exception, it is necessary to develop a plan for handling medications supplied by patients. Although we have a hospital policy that strongly discourages the use of patient-supplied medications, the situation is occasionally unavoidable, necessitating that a policy be in place. Our current process requires that we package a supply of the patient’s medication in unit dose. We generally use the manufacturer’s bar code on the package if it is included on the prescription label or can be determined. In cases when that is impossible, we package using an order-specific bar code on each unit dose (ie, the same bar code that is used on our compounded items).
A Team Effort
Collaboration among departments contributed greatly to our overall success, and from the beginning, pharmacy and nursing worked together well. A committee of leaders and staff from both departments met regularly to identify opportunities for improvement and build action plans to resolve issues. Expanding this group to include additional bedside staff nurses, as well as information technology (IT) staff, was critical to the successful roll-out of BCMA. To their credit, every decision concerning the system—from determining the type of bar codes to put on wristbands to selecting the instances for which scanning would not be required—was made collaboratively.
This committee also was very hands-on; both pharmacy and nursing trialed scanners and determined the set-up for computers in medication rooms, as well as mobile stations in patient rooms. Nurses completed over 300 real-time observations of medication administrations to identify any gaps in adherence to policy that would require addressing before moving forward, and pharmacists provided key input into flow-charting both the former medication administration workflow and the anticipated workflow after BCMA implementation. When it came time to build the computerized documentation that would support BCMA, this interdisciplinary group determined every aspect, including administration actions, override reasons, and partial dosage documentation. In addition, pharmacists were included in the training programs that were offered to bedside nurses because we believed it was important to the overall success of the project that pharmacists and nurses learn the content together.
At the outset of the implementation, the medical/surgical units adopted BCMA at a rate of one unit per week. After we felt comfortable that the majority of system and medication issues were resolved in these first units, we began implementing on multiple units simultaneously. However, with the specialty areas (ie, critical care, pediatrics, etc), we reverted to the one-unit-at-a-time strategy.
The installation process was the same for each unit: the week prior, nursing and pharmacy leads met with unit management to review the schedule, describe the role of super-users (ie, provide support to users and troubleshoot and resolve issues quickly), and set expectations. The leads instructed nurses to halt administration if they received an alert indicating either that the patient or medication bar code was not recognized or that the patient did not have an active order for the scanned medication. If that occurred, they were to contact a super-user to review the situation and devise a plan for proceeding.
The super-users found that, in many instances, the alerts were due to end-user errors that were remedied easily with one-on-one education. In some instances, bar codes would scan, but were not recognized as a medication for which the patient had an order. This generally happened with over-the-counter products that had multiple, but unlinked, medication records in the EHR database. For example, we purchased a zinc oxide ointment, but the EHR orders were linked to a zinc oxide paste. This caused the system to fire an alert of no order for the scanned medication. We also had some medications with bar codes that simply were not recognized. Frequently these were the result of having stock in an ADC that had not gone through the validation process.
An unexpected circumstance developed when nurses used their clinical judgment to administer only a portion of a prescribed pain medication; the system issued an alert indicating the patient did not have an order for that dose. Also, there were instances when a nurse administered a medication that was prescribed as a one-time dose—our system assigns a time window of 1 hour for these orders, so when the nurse attempted to scan and administer the medication after the hour had passed, an alert was generated indicating the patient did not have an active order. To fix these situations, we had physician leaders verify the intent of these orders and had the IT build team make changes to the system in how the orders appeared to the end user. By having a nurse and pharmacist super-user available on the unit, these situations were identified and resolved quickly. Although resolution of issues occasionally required input from medical administration, providing education to end-users was sufficient most of the time.
Despite our comprehensive preparation for BCMA implementation, the challenge of achieving high scan rates became quickly apparent; although we were scanning almost 98.8% of medications and just over 99% of patients (see FIGURES 1 and 2), this still meant 7000 medications per month were escaping the scanning process. Our quality improvement executive director encourages us to accept nothing less than 100% compliance for patients and medications for which scanning is possible and expected. Thus, we explored a number of strategies to determine the causes for non-scans.
Using Data to Improve Scan Rates
The IT department built a number of reports to be available immediately upon implementation of the new BCMA system, including medication- and patient-scanning compliance reports, near-miss reports, and medication administration record edit reports. Each afternoon during an implementation, unit meetings were held to address issues and concerns. On days two through four of the process, the meeting included a review of the compliance reports from the previous day. These reports were compiled daily and listed instances in which a patient or medication was not scanned. They included data from the scanners themselves, to help identify issues with medication and wristband bar codes. Sharing compliance information at this early stage was based on the premise of inspect what you expect. After reviewing the compliance reports, nursing leaders discussed the findings with the staff involved in an effort to identify real issues with the system versus simple non-compliance. Leaders who reviewed these reports and talked to staff regularly had the greatest improvement in their units’ overall scan rates. Still, compliance, while good, remained fairly stagnant.
Near-miss reports, which are primarily reviewed by pharmacy, list every time a scanned medication signals an alert to a nurse that the drug was not ordered for the patient. Also, they list administrations that are manually charted after an alert fired. The ability to deliver this information to unit management in almost real time enabled pharmacy to provide encouragement and education to users who were struggling or electing not to scan because they believed the system was of questionable value. The reports also created a means by which to determine the effectiveness of the BCMA system, as they established the number of medication errors being prevented by BCMA. Every month, approximately 5000 alerts fired indicating either that a medication order was discontinued or that the order was not active for the patient. Of these alerts, approximately 500 were associated with subsequent manual documentation of a medication administration. However, not all of these documentations indicated a medication error. In fact, just a small percentage of administrations were actual medication errors. BCMA reduced wrong medications to patients 41% in the first year after implementation (see FIGURE 3). As more units went live with BCMA, errors were reduced further.
The team also relied on a report that listed every time an edit of any kind was made to the default medication administration variables (ie, a change to the time or date of the scan or dosage of the medication scanned). This report revealed whether medications were actually scanned prior to administration or if nurses scanned the medication and then changed the time of administration to precede the scan time. One of the unintended consequences of reviewing compliance reports and addressing non-scans is a shift from no scanning to back scanning (scanning after a medication has been administered). Several self-reported medication errors have resulted from scanning after medication administration and then being alerted that the medication was not ordered for the patient. We have used these events to provide supplemental education to nurses specifically to eliminate the use of back scanning as a legitimate medication administration process.
Reviewing reports and holding staff accountable for deviations from established practice standards, while critically important, can only improve scan rates to a certain extent. To bolster our efforts, we looked to the Kaizen methodology to guide us in determining what more could be done to increase the frequency of successful scans. We instituted monthly BCMA Kaizen activities, with both nurses and pharmacists, to review real-time reports of non-scans and medication administration record edits. To date, eight events comprising 281 observations have transpired. The distribution of observations is shown in FIGURE 4. Observations were organized into nine categories (see TABLE 1), with human factors accounting for nearly half of the reasons for non-scans or edits to the medication record. In response, we removed the requirement to scan certain products. For example, the removal of a patch from a patient previously required a scan, but there was essentially no product to scan. Likewise, clove oil was documented on the medication record, but was being used as a room freshener to be kept at the patient bedside.
Kaizen activities also unearthed variations in nursing practice related to medication administration, as well as in knowledge regarding some aspects of the scanning procedure. As a result, a medication safety survey was sent to all of the nurses working on units with live BCMA scanning. The survey was voluntary, but over 400 nurses completed it. The nurses were asked to identify what BCMA confirms. Between 50% and 60% of the nurses thought scanning bar codes of medications had the ability to alert them to incorrect route and times and that scanning took the place of patient identification; 91% knew that BCMA scanning verified that the correct medication was ordered for the patient.
In addition, nurses were asked to identify workarounds that they either witnessed or engaged in themselves. The results are shown in FIGURE 5. Three of the workarounds relate directly to the scanning process and included scanning medications after administering them, scanning an IV bag already hanging versus a new one, and not scanning a medication. This reinforced the need to not simply monitor compliance reports, but rather to improve systems to make it easier for nurses to scan consistently.
We know that the collaborative work of pharmacy and nursing involving BCMA does not stop here. The system requires ongoing vigilance to ensure that all medications and patients in the facility can be, and are, scanned prior to medication administration, and that the system optimally facilitates scanning in all circumstances so users are not enticed by workarounds to bypass a sure route to patient safety.
Stephanie A. Rybak, PharmD, is senior assistant director of pharmacy for operations and supply chain at Beaumont Health System–Royal Oak in Royal Oak, Michigan. For nearly 15 of her 30 years at Beaumont, Stephanie was the assistant director for women’s and children’s health. After that role, she became assistant director of pharmacy technology. She was the pharmacy lead for the BCMA implementation. Stephanie received her PharmD from the University of Nebraska College of Pharmacy.
Anne Ronk, RN, MSN, NEA-BC, has been the nursing director of women’s, children’s, and psychiatric services at Beaumont Health System–Royal Oak for 20 years. After graduating from the nursing school at University of Michigan with her bachelor’s degree. Anne began her career as a staff nurse in the pediatric intensive care unit. She then obtained her MS in nursing care of children and adolescents from Wayne State University in Detroit. She is certified in nursing administration and has been a staff development instructor and a clinical nurse specialist.
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.