NU 725 Unit 4 systems evaluation example

Evaluating Barcode Medication Administration as Configured on a 30-Bed Surgical Unit

Student Name

Department of Nursing, Herzing University

NU 725: Technology and Nursing Informatics in Advanced Practice

Instructor Name

February 10, 2026

Evaluating Barcode Medication Administration as Configured on a 30-Bed Surgical Unit

Criteria, Committed to Before Looking

The system under review is barcode medication administration (BCMA) as installed on a 30-bed composite surgical unit, where nurses scan the patient's wristband and each medication before administration and the system checks the scan against the active order. Six criteria were set before any data were reviewed, organized loosely around the dimensions of information systems success proposed by DeLone and McLean (2003): system quality, information quality, use, user satisfaction, and net benefits. Each criterion has a definition of what would count as meeting it.

Criterion 1, scanning compliance: at least 95% of administered doses have both the patient and the medication scanned. Criterion 2, error interception: the system produces actionable warnings that stop wrong-patient, wrong-drug, and wrong-dose events, with evidence from the unit's warning log. Criterion 3, override behavior: fewer than 10% of warnings are overridden, and override reasons are documented. Criterion 4, workflow burden: routine administration of a single oral medication takes no more than five discrete steps, and the process does not require workarounds. Criterion 5, information quality: the medication administration record matches what was given, with time stamps accurate to within five minutes. Criterion 6, user acceptance: nurses report that the system helps rather than hinders safe administration.

Evidence Under Each Criterion

Criterion 1, compliance. The system's compliance report for the past three months shows patient wristband scanning at 97% and medication scanning at 89%. Medications not scanned were concentrated in three groups: intravenous piggyback bags, whose labels printed by pharmacy often fail to scan; liquid oral medications drawn up from bulk bottles; and doses given during rapid response events. Verdict: not met for medication scanning.

Criterion 2, interception. The warning log recorded 412 warnings over the three months. A review of a random sample of 60 found 9 that prevented a real error, including 2 wrong-patient scans and 4 doses given too early. That matches published evidence: when a large academic hospital compared units before and after adoption, barcode technology combined with an electronic administration record reduced administration errors unrelated to timing by about 41% and potential adverse drug events by about half (Poon et al., 2010). Verdict: met, with evidence of real interceptions.

Criterion 3, overrides. 138 of 412 warnings, 33%, were overridden. Most overrides were for timing warnings on scheduled doses given within the hospital's allowed window, which the system flags because its timing tolerance is set more narrowly than policy. Verdict: not met, largely because of a configuration choice.

Workflow, Workarounds, and Alert Burden

Criterion 4, burden. Observation of 20 routine oral administrations counted a median of seven discrete steps, including logging in, opening the record, scanning the wristband, scanning each medication, reviewing a pain reassessment prompt, confirming, and signing. Two workarounds were observed: a nurse carrying pre-printed spare wristband labels for a patient whose band would not scan, and another scanning a medication's barcode from the dispensing cabinet drawer rather than the dose itself. Both are well-described workarounds that defeat the safety check they bypass (Koppel et al., 2008). Verdict: not met.

The override rate and the timing warnings are linked. When more than a third of warnings do not require action, nurses learn to click through them, and the warnings that matter lose force. Alert burden here is not an abstract risk; it is visible in the override pattern.

Information Quality and User Acceptance

Criterion 5, information quality. A comparison of 40 administration records with direct observation found the recorded time within five minutes of the actual time in 36, or 90%. The four discrepancies were doses documented after the fact at the end of a medication pass. Verdict: largely met.

Criterion 6, acceptance. A short survey of 22 nurses on the unit found that 18 agreed that the system catches errors they would otherwise miss, but 15 said it slows them down, and 12 said the timing warnings are "mostly noise." Verdict: partly met; nurses value the safety function and distrust the warnings.

Limits of This Evaluation

Three limits affect how far these verdicts can be trusted. First, the warning log sample of 60 was small, and the count of nine real interceptions could easily be several higher or lower across the full log; the verdict on criterion 2 rests as much on the published evidence as on the local sample. Second, the observations of workflow were done on day shifts, when staffing is highest, and may understate the burden on nights, when one nurse may cover more patients and workarounds are more tempting. Third, the survey reached 22 of the unit's 38 nurses, and those who responded may differ from those who did not.

None of these limits reverses a verdict, but they shape the recommendations. The workflow measures should be repeated on nights before the pain reassessment prompt is moved, and the survey should be repeated after the changes, with a reminder sent to every nurse on the unit, so that the before-and-after comparison uses the same population.

Overall Judgment

The system meets its core purpose: it intercepts real errors, and nurses recognize that. It fails on three criteria, and in each case the cause is local rather than inherent to the product: a pharmacy label printer that produces unscannable barcodes, a timing tolerance set narrower than hospital policy, and a sequence of prompts that adds steps without adding safety. The unit is therefore getting part of the safety benefit reported in the literature and paying a higher workflow cost than necessary.

Recommendations

First, replace or recalibrate the pharmacy label printers and track the intravenous piggyback scan rate weekly until it exceeds 95%. Second, widen the timing tolerance to match the hospital's medication administration policy, and measure the override rate again after one month, with a target below 15%. Third, move the pain reassessment prompt out of the administration sequence and into a separate task, reducing the routine steps from seven to five. Fourth, stock bulk liquids as unit doses where possible. Each recommendation targets a specific failed criterion, and each has a measure that can show whether it worked.

References

DeLone, W. H., & McLean, E. R. (2003). The DeLone and McLean model of information systems success: A ten-year update. Journal of Management Information Systems, 19(4), 9-30. https://doi.org/10.1080/07421222.2003.11045748

Koppel, R., Wetterneck, T., Telles, J. L., & Karsh, B.-T. (2008). Workarounds to barcode medication administration systems: Their occurrences, causes, and threats to patient safety. Journal of the American Medical Informatics Association, 15(4), 408-423. https://doi.org/10.1197/jamia.M2616

Poon, E. G., Keohane, C. A., Yoon, C. S., Ditmore, M., Bane, A., Levtzion-Korach, O., Moniz, T., Rothschild, J. M., Kachalia, A. B., Hayes, J., Churchill, W. W., Lipsitz, S., Whittemore, A. D., Bates, D. W., & Gandhi, T. K. (2010). Effect of bar-code technology on the safety of medication administration. New England Journal of Medicine, 362(18), 1698-1707. https://doi.org/10.1056/NEJMsa0907115