Medication errors cut by bar-coding and observation

Methodology lends itself to measurement

A 54% reduction in medication administration errors was recorded at Pennsylvania's Lancaster General Hospital as a result of direct-observation methodology used to monitor medication administration before and after deployment of electronic medication administration records and bar-coded medication administration.

In a research study reported in the American Journal of Health-System Pharmacy, Richard Paoletti, MBA, BSPharm, and colleagues say the goals for a multidisciplinary approach to systematically decrease medication errors through use of observation methodology and deployment of electronic medication administration records (EMAR) and bar-coded medication administration (BCMA) included: 1) effectively using an observation-based medication error detection system to provide a reliable and effective mechanism for identifying medication errors during administration; 2) identifying and implementing a best-practice technology solution for EMAR and BCMA verification; 3) improving pharmacy and nursing communication; 4) maintaining nursing satisfaction through minimizing additional workload; 5) projecting a patient safety leadership role to the community; and 6) establishing a data source to allow for continual and effective identification of systems issues to facilitate a proactive approach to prevent medication error occurrence.

Joining Lancaster General Hospital's multidisciplinary team for this project were representatives from the administrative, medical, nursing, pharmacy, information services, security, marketing, registration, and staff development departments. The team was formed to improve the safety of the medication administration process through a reduction in medication errors.

Direct observation facilitates measurement

The team decided to use a direct observation process to get an accurate measurement of current error rates, system enhancements, and goal achievements because that method allows hospitals to identify faulty processes, recommend system changes, and evaluate changes with consistent and reliable data. The observation technique was used in addition to the hospital's self-reporting system rather than in place of it.

Four nurses were trained in a nationally recognized certification program for the observation process. These four certified medical observers observed nurses throughout the entire medication administration process during peak workload periods in selected nursing units. Bedside observations were recorded and later compared with patients' charts to identify discrepancies between observed medication administration and physician orders.

Three inpatient nursing units participated in the study—intervention group 1 was a 20-bed cardiac telemetry unit and intervention group 2 was a 36-bed medical-surgical unit, while a 20-bed cardiac telemetry unit was the control group and was included to monitor for other variables that may affect the medication administration process.

During the study's first phase, the three units participated in evaluating the medication administration process with a five-day medication administration record. The observers noted that intervention group 1 had a lower variation in medication orders because of the aggregated cardiac patients and use of standardized order sets. That group had also implemented a double-check process using two nurses to verify transcription of all medication orders. Because of those variances, the observers believed the medication administration process for intervention group 2 was a more valid comparator with the control group.

The observers found 188 total errors. The vast majority were classified as wrong time (143) and were typically late doses. Others types of errors noted included omission (15), wrong technique (14), wrong dose (6), extra dose (5), wrong medication (3), wrong route (1), and wrong formulation (1).

Opportunity for significant improvement

Extrapolating those findings to the hospital's 521 licensed beds, which are at more than 90% capacity, the team saw an opportunity to make a significant improvement. They quickly identified the hospital's handwritten, non-pharmacy-generated, five-day medication administration record as a problem area. They also found that the administration phase of the medication-use process was particularly vulnerable since safety nets relied on nurses to remember, identify, and resolve discrepancies at the bedside. A decision was made to implement a pharmacy-driven EMAR and BCMA system.

Issues addressed before implementation of the new system included:

Infrastructure. A wireless network was necessary for nurse workflow and future clinical information system needs.

Device and stand selection. Staff nurses were asked to participate in selecting the devices and stands, including the quality and availability of equipment. A mobile computing tablet with a bar-code scanner was selected, with consideration for nurse workflow related to medication administration and future clinical documentation needs. One device per staff nurse was bought, along with an additional device per nursing unit.

Pharmacy. Pharmacist order-entry activities lacked much of the clinical parameters included in physician orders. Considering the move to a pharmacy-driven EMAR, an extensive amount of work occurred in the pharmacy relating to completeness of order entry. Staff performed several system enhancements to the clinical information system to comply with EMAR specifications. Also, action orders were created in the pharmacy system to prompt nurses with clinical notes and monitoring parameters for specific medications. A paper, fax-based nursing-pharmacy tool was developed to enhance communication of order-entry issues. Barcoded packaging systems were investigated and used.

Different labeling formats and packaging techniques were piloted, with a focus on quality to provide nurse end-users with accurate and functional bar codes. Pharmacy labels were reformatted to accommodate the bar codes. Staff say they gave careful consideration to bar-code content, and most bar-codes were derived from the product National Drug Code. Codes for some specialized products such as pediatric preparations and compounded solutions were based on the order number generated by the pharmacy information system.

Medication administration guidelines. All workflow issues involved in the medication-use process were identified through nurse-pharmacist collaboration. Modifications made because of the effect of EMAR and BCMA included nursing-pharmacy communication, real-time reconciliation of medications ordered, administration of medications, and documentation strategies.

Communication. A plan was established to communicate this patient safety initiative to employees, physicians, patients, families, and community members. The team created a patient handout to explain the intention to provide patients with a safer medication administration environment.

Employee identification. A process was established to change the employee identification badges to accommodate a bar code for access to the BCMA system.

Patient identification. Paoletti says the first step in "five-rights" verification is patient identification. To ensure patients are identified correctly, the team determined that all patients needed to have bar codes on their wristbands. That necessitated a new system implementation and consideration of the specific needs of the BCMA system including quality, durability, and security of the wristbands. The team mandated that the wristbands could only be obtained through the admissions office to protect system integrity. Nurses were not granted access or permitted to reproduce the wristbands as part of the plan to minimize unsafe work-arounds. Ongoing integrity of the BCMA system for patient safety continues to be a priority in projects that use barcode technology. Thus, when bar-coded chart labels were introduced to replace an existing patient demographic stamp system, the chart label was created without the ability to be scanned to ensure use of the patient wristband in the BCMA system.

User education and support planning. The hospital recognized that implementing EMAR and BCMA presented nurses with a significant practice change in terms of the medication administration process. A four-hour class and additional computer-based learning modules were provided for all nursing personnel. Nurses also observed the process on units using EMAR and BCMA before their implementation. Support was provided on each nursing unit 24 hours a day for the first seven days of "go live" implementation. Also, pharmacists and physicians were encouraged to participate in the training program to gain an understanding and appreciation of the effect the innovative system would have on nursing.

Downtime process. A contingency plan was created and established to cope with scheduled or unplanned downtime.

Additional personnel. Paoletti says appointment of a dedicated clinical nurse as system administrator was a significant key to the initiative's success. Responsibilities for this position included oversight and daily system monitoring, data analysis and integrity, and establishing useful reporting. Nurses typically reported a slight increase in time needed to support medication administration because of the scanning process, but this factor has not been scientifically measured. The need to repackage and bar code a substantial amount of medication resulted in addition of one pharmacy technician dedicated to that task. And the team says an additional pharmacist for checking, validating, and managing the repackaging effort would be a valid consideration.

The new EMAR and BCMA systems were implemented on one nursing unit beginning August 2003 and spread to all patient units by July 2004. The team originally wanted to have hospital-wide implementation completed by December 2003 but the date was pushed back to accommodate training demands.

Competing systems make transfers difficult

During the implementation period, internal patient transfers became difficult because of the difference in systems on various nursing units. The planned implementation scheme was revised to deploy "feeder" units in sequence. An additional obstacle that was uncovered was the unavailability of hallway electrical outlets because of the hospital's age. The facilities department had to add additional power sources to many of the nursing units and extended-life batteries were obtained to support the new equipment.

In the study's second phase, the control group continued to use the five-day MAR without changing the process, while the two intervention groups were measured to evaluate the medication administration process using EMAR and BCMA technology. The intervention groups were monitored to allow for total visibility to different practices affecting the medication delivery process.

Paoletti reports that post-implementation data for intervention group 2 demonstrated a 54% reduction in medication errors. While the same reduction was not noted in intervention group 1, the observers identified differences in practices during the study's baseline phase in comparison with the control group.

Now in all impatient units and other areas

Today, the EMAR and BMCA program is deployed in all inpatient nursing units and in additional care areas including the medical outpatient unit, outpatient infusion center, cardiac catheterization laboratory, and post-anesthesia care unit. Paoletti says implementation of the new system "has provided significant enhancements to medication administration. In addition to the noted reductions in medication error rates, this program has enhanced our ability to capture valuable data related to medication administration and compliance with established medication administration processes."

One of the most important aspects of this medication safety initiative is the ongoing systematic, direct observation of medication administration on nursing units, Paoletti says. During unannounced, regularly scheduled observation, the certified observers continue to monitor the entire medication delivery process, allowing for early identification of non-compliance to established procedures. He says the technique also provides accurate information to decision-makers who are better equipped to make procedural changes to continually enhance medication-use system safety.

The direct observation accuracy rate before BCMA was 86.5%. After implementation of the bar-code system, the accuracy rate rose to 97%. Paoletti says the system has led to a 54% decrease in the medication administration error rate.