Improving the Medication Administration Process: The Impact of Point of Care Bar Code Medication Scanning Technology

Steve Rough, M.S., R.Ph., Brad Ludwig, M.S., R.Ph., Edie Wilson, R.N., B.S.N.

University of Wisconsin Hospital and Clinics

Madison, WI

Literature has demonstrated that over 30% of preventable adverse drug events occur at the administration phase of the medication use process. The use of bar code technology can improve the accuracy of medication administration and documentation resulting in safer patient care. In December of 2001, the University of Wisconsin Hospital and Clinics implemented a handheld wireless point of care medication bar code scanning system (BCSS) used to verify and document medication administration at the patient’s bedside on 28-bed inpatient pilot unit. The system has since been expanded to 22 of 26 inpatient care areas. This was the third part of an integrated strategy to apply advanced automation technology to enhance medication safety, essentially closing the loop of the medication use process.

The primary outcomes of this project were to improve medication administration and documentation accuracy and improve nursing satisfaction with the medication use process. Specific goals of the project included:

1. Implement BCSS for inpatient medication administration and documentation on all inpatient nursing units.
2. Measure the impact of this technology on medication administration and documentation error rates.
3. Assess the satisfaction of nurses with the new medication administration and documentation system versus the prior paper-based medication administration documentation system (kardex).

The project was coordinated by two pharmacy managers and one nursing clinical director. Implementation of BCSS required input from many different disciplines including pharmacy, nursing, information systems, admissions, and respiratory therapy. Eight multidisciplinary teams worked on project implementation.

Nurse satisfaction with the traditional paper-based kardex system was measured four months prior to implementation using an eleven-element, 1-to-5 likert scale instrument. These results were compared with a six-month post-implementation survey using a twenty-element, 1-to-5 likert scale instrument to collect data.

Baseline medication administration and documentation error rates were collected using a direct observation technique, in which the trained observer was blind to the patient’s medical history. The observers utilized a data collection form to standardize the collection of medication administration elements. After each observation was complete, the patient’s medical record was audited. The audit compared the medication-related activities of the nurse at the time of medication administration with the actual physician order. A total of 450 observations were recorded on the pilot unit prior to BCSS implementation, with 1,245 observations conducted organization wide for pre-implementation evaluation. Data collected post-implementation utilized a series of vendor created reports that provided a list of all errors recorded within the BCSS software, including all warning messages generated by BCSS for a designated period of time, and whether the nurse heeded the warning. All actual and potential medication administration errors were captured electronically at the patient’s bedside during this audit period. A medical chart audit occurred after each report to compare the medication related activities of the nurse at the time of medication administration with the actual physician order. These data were collected over a seventeen-day period two months after implementation, netting 7,013 doses checked by the BCSS software on the pilot unit.

Implementation of the BCSS has dramatically reduced medication administration error rates and improved nurse satisfaction with the medication administration and documentation system. The overall pre-BCSS medication administration error rate was 9.09% with the most common error being wrong time (3.96%, or 44% of total errors). The overall post-BCSS medication administration error rate was 1.21%. While all categories of error types were improved, omitted dose errors demonstrated the largest improvement declining by 92% from 1.4% to 0.11%, while wrong time errors were reduced by 77%. Improper dose and wrong dosage form were eliminated as an error source.

The BCSS warning reports allowed for the determination of the number of potential errors or near-misses that BCSS was able to intercept at the bedside allowing the nurse to correct the situation prior to administering the medication. These potential errors would likely have reached the patient had they not been intercepted by BCSS. 3.2% of doses scanned on the pilot unit were intercepted as a near-miss; this translates to 4,657 doses annually on the pilot unit.

Analysis of occurrence report data before and after BCSS implementation on the pilot unit demonstrates that reported medication administration errors via the hospital’s non-punitive reporting system have decreased by 79%. However, overall error rates have remained consistent. Thus, other medication errors have been increasingly reported post-BCSS implementation. In particular, errors involving IV pump programming, pharmacy order entry, and prescribing near-misses have become more prevalent post-implementation.

Comparison of nursing satisfaction data demonstrated a 42% improvement in overall nursing satisfaction with the medication administration and documentation system after BCSS implementation. Initially some nurse resistance results due to the fact that scanning the patient, medication, and nurse added some additional time to the administration process. However, this resistance was quickly overcome as staff saw errors prevented by the system.

Despite the dramatic improvement in medication administration accuracy and nurse satisfaction, BCSS technology introduces new sources of error that are currently being analyzed and managed. There are several projects under way to continuously improve the medication administration process. A few opportunities for continued improvement in the BCSS system include advanced software development to make the system more intuitive for nurses, product reporting capabilities to help providers easily identify trends in past doses administered and integration of bar coding capability with new IV “smart pump” technology.

Medication error reduction requires a comprehensive approach to identify and eliminate error potential. The administration phase of the medication use process appears particularly vulnerable, since safety nets currently rely on humans to remember, identify and resolve problems at the bedside. This is often difficult, since the bedside is a stressful environment that lacks information support. In order to reduce medication administration errors, the implementation BCSS for medication administration was pursued. Medication administration errors have been significantly reduced as a result of BCSS. Nurses are satisfied with the BCSS system and report improved care planning and improved information access with use of the device.