Using Data Mining to Identify and Reduce Problematic Bar Codes

July 2010 - Vol. 7 No. 7 - Page #10

By Ron Schneider, Jonathan Bagby, and Elizabeth Mims

The Department of Veterans Affairs (VA) implemented bar code medication administration (BCMA) throughout its medical centers in 1999—a time when few pharmaceutical manufacturers applied bar codes at the unit-of-use level. In order to enable this process, VA facilities had to manually print and apply locally generated bar codes to its unit-of-use packaging. Each facility was responsible for ensuring the bar coded medications were scannable prior to dispensing at the point of care. At that time, no national oversight existed within the VA health care system—known as the Veterans Health Administration (VHA)—to provide field guidance or troubleshoot when medications failed to scan.

Early Attempts to Address BCMA Errors
In order to track and troubleshoot non-scanning medications, individual VA health care facilities created their own manual reporting mechanisms for BCMA end-users to notify the pharmacy of failed scans. These rudimentary reporting processes, such as returning the medication to the pharmacy in a designated “failed-to-scan” drawer on the medication cart or sending the actual bar code unit dose packaging (minus the medication) to the pharmacy, were labor intensive, requiring pharmacies to locate the actual medication package reported and then re-label it. While these mechanisms helped individual VHA medical center sites identify and replace problematic bar codes one medication at a time, they failed to address the system’s latent issues—such as bar codes not being properly applied, bar code information not appearing in the data dictionary, or poorly printed bar codes.

To address these and other issues related to BCMA, the VA’s national Bar Code Resource Office (BCRO) was established in 2003. One of the first programs created by the BCRO was the Clinical Closed Loop Verification Program for VA medical centers to report problematic bar codes.1, 2 Verification is the bridge between the print quality of the bar code and the successful scan rate of the bar code. Global Standards 1 determines bar code quality using “a special piece of equipment called a barcode verifier ... The verifier takes nine precise quality measurements that cover various aspects of bar code print quality. The overall quality is expressed in either a numerical or a letter grade ranging from zero (F) to 4.0 (A). The higher the grade, the more likely that the symbol will scan successfully.”3 VHA has set standards for its bar codes to scan at least a C grade with a target of A.

Once we had a system in place to identify and grade bar codes, we began working on a system to organize and apply this data. In 2006, VHA Directive 2006-0081, Quality Directive for Unit Dose Packaging and Bar Code Labeling, established a quality control system to ensure bar codes scanned properly prior to leaving the pharmacy. End-user direct observations during medication passes also were conducted to validate that products dispensed from the pharmacy scanned properly. Although this process created additional workload for the pharmacy, it provided a safety net for the BCMA application by reducing the number of workarounds made by nursing for medications that did not scan.

Collecting and Using Failure Data
Once a system to track bar code validation was in place, facility pharmacies began reporting their results quarterly to the BCRO via a manual reporting system. The initial baseline data reported in March 2006 showed an overall medication scan success rate of 95%. While this rate may seem impressive initially, consider that approximately 678,000 medications were being administered daily, thus the number of failed bar codes came to approximately 34,000 per day, and each failed scan represented an opportunity for error due to scanning circumvention. Fortunately, by this point we had established a formalized process through the Clinical Closed Loop Barcode Verification Lab—which reported failed scans to the BCRO—to identify the failure points of the bar code.

If the problem is truly a print-verification problem, the BCRO will contact the supplier, notify them of the verification results and failure points of the bar code, and request that they make improvements. Noted problems with a manufacturer’s bar codes are also reported through the Food and Drug Administration (FDA) MedWatch Program, unless exempted (eg, OTC products, nebulizers, and polyvinyl polymers) from the February 26, 2004 ruling on Bar Code Label Requirements for Human Drug Products and Biological Products (69 FR 9120). Conversely, if the problem lies with a bar code generated at a medical center, the BCRO communicates both the verification results and the failure points to the medical center and provides guidance on how to address the identified problem.

Real-time Failure Notification
In March 2009, the VA developed an enhancement to its BCMA software, called Managing Scanning Failures (MSF), to provide just-in-time reporting of bar codes on medications and patient wristbands that do not scan to all VA medical centers. Using this new software functionality, when the bar code does not scan successfully, the nurse clicks the “Unable to Scan” button on the screen and selects the believed reason why the medication does not scan from a drop-down list (eg, damaged wristband, equipment failure, etc); there are three options for wristbands and five for medications. The nurse proceeds with the medication administration after verifying the five rights and typing the manufacturer’s national drug code (NDC) from the package into the system. The “Unable to Scan” function automatically generates an e-mail message to notify selected pharmacy personnel to rectify the issue before the next medication administration time. Lastly, the nurse visually confirms the medication, dosage, and dosage form.

Data Mining
Subsequent to the release of MSF, an automated data mining process was developed. The process allows for a national data extraction each month, and makes that data accessible to VA offices through a Web-based reporting tool (see Table 1).


The report includes a summary of drugs reported as unable to scan. The “Medication Unable to Scan–Drug Listing” displays the medications in descending order by number of “unable to scan” reports. In this case, there were 2723 instances of doses of Insulin Regular Human 100 unit/mL reported as “unable to scan” across all VA medical centers during a 30-day period.

The individual drug names are hyperlinked to secondary tables containing the facilities reporting each drug as unable to scan, sorted by order of frequency (see Table 2). This one-click filter allows BCRO staff to quickly identify those facilities reporting the highest incidence of specific drugs with scan failures, and offer assistance with corrective actions.


The initial baseline data of VHA Directive 2006-0081 reported in March 2006 showed an overall medication scan success rate of 95%, or failed scan rate of approximately 34,000 medications per day. By May 2009, the medication scan rate had increased to 99%, or a failed scan rate of approximately 6780 medications per day. When viewing manufacturer labeled products only, the number of failed scans per day was reduced to approximately 4700 mis-scans.

Bar coded medication administration has proven its effectiveness in helping prevent medication errors, but it is also a system that requires vigilance and maintenance. Performing data mining of real-time information on bar code failures has given us the ability to address and fix those problems quickly, and by providing feedback on our data to drug manufacturers, this functionality helps the entire medication distribution system. Performing meaningful analysis of your BCMA system will not only improve your own effectiveness, but by providing feedback, you too can affect wide reaching improvement to this important element of health-system pharmacy operations.


  1. Department of Veterans Affairs. VHA Directive 2006-008. Quality
  2. directive for unit-dose packaging and barcode labeling. February 9, 2006.
  3. Schneider R, Bagby J, et al. Resolving bar code issues in a BCMA system. Pharm Purch Prod. 2009;6(10): 2-6.
  4. Barcode Quality. Global Standards 1. Accessed June 17, 2010.

Ronald Schneider, RPh, MHA, is the pharmacist consultant for the Department of Veterans’ Affairs Bar Code Resource Office and has an extensive background in pharmacy automation systems and operations. Ron holds a BS in pharmacy and a master’s in health care administration from the University of Maryland.

Jonathan Bagby, MSN, MBA, RN-BC, is a nurse consultant for the Department of Veterans’ Affairs Bar Code Resource Office and has a background in medical, surgical, and orthopedic nursing. Jonathan holds a BS in nursing from Virginia Commonwealth University and a master’s of science in nursing and business administration from the University of Phoenix.
Elizabeth A. Mims, MBA, BSN, PMP, is deputy director of the Department of Veterans’ Affairs, Bar Code Resource Office. Elizabeth holds a BS in Nursing and a Master’s in Business Administration from the University of Phoenix.


Like what you've read? Please log in or create a free account to enjoy more of what has to offer.

Current Issue

Enter our Sweepstakes now for your chance to win the following prizes:

Just answer the following quick question for your chance to win:

To continue, you must either login or register: