Accurate medication administration documentation underpins patient safety, legal record-keeping and continuity of care across clinical settings. Electronic medication management systems offer advantages over paper charts, including greater legibility, remote access, clinical decision support and automated documentation fields. However, the accuracy of medication administration records remains a concern when interfaces are confusing or automated fields are not checked. A stepped-wedge cluster randomised trial published in BMJ Health Care Informatics compared medication administration documentation with paper charts and an electronic medication management module in a tertiary paediatric hospital in Sydney, Australia. The comparison focused on completeness, accuracy and compliance with legal requirements during the transition from paper-based medication charts to electronic records.
Medication Documentation After Digital Transition
The electronic medication management module formed part of an existing clinical information system and supported electronic prescribing, recording of drug dispensing, drug administration, medication reconciliation and monitoring. Its functions included predefined order sentences, age-based and weight-based filtering, a dose calculator, complex titratable infusions, order sets, nurse-administered medications and allergy and drug-drug interaction checking.
Medication administration processes were observed across nine general wards over 10 weeks. At the start, all wards used paper medication charts. Wards moved to the electronic system at one-week intervals, with all wards using electronic medication management after eight weeks. Trained nurse observers followed randomly selected nurses who had already consented to take part and recorded administration details with the Precise Observation System for the Safe Use of Medicines.
The analysis included 5095 dose administrations after excluding 42 wrong-drug administrations, because other errors were not assessed for those cases. The included administrations involved 1558 patients and 298 nurses. Patient, nurse and administration characteristics were broadly similar across paper and electronic periods, although patients were older during the electronic period because of ward populations and the order of implementation.
Completeness Improves as Accuracy Falls
Electronic medication management improved some areas of documentation completeness. The proportion of administrations without a nurse signature did not differ significantly between paper charts and electronic orders, at 1.3% and 1.0% respectively. Missing administration times were significantly less common with electronic medication management than with paper charts. The adjusted odds ratio for no documented time was 0.15 with electronic records compared with paper.
The strongest concern involved inaccurate documentation of what actually happened during administration. When electronic medication management was used, documented administration details were more likely to differ from what trained observers saw. The adjusted odds ratio for observed administration differing from the record was 4.6. Examples included records showing an intravenous bolus when the observed administration was an intravenous infusion, or the reverse.
Electronic records were associated with fewer doses administered from incorrect, illegal or incomplete prescriptions. The adjusted odds ratio was 0.25 with electronic medication management compared with paper charts. However, conflicting instructions were more common with the electronic system, with an adjusted odds ratio of 2.3. One example involved a prescriber ordering an intravenous bolus while ancillary information in the system specified an intravenous infusion.
Workflow Patterns Affect Recorded Times
Signing before administration appeared common under both systems, even though policy required a different sequence. Nurses signed before giving the medicine in 94.1% of paper-chart administrations and 86.6% of electronic medication management administrations. Signing most often occurred within five minutes before the drug was given, suggesting a workflow in which nurses signed the chart and then administered the dose shortly afterwards.
In a quarter of cases, nurses signed more than five minutes before giving the medication. This occurred in 25.3% of paper-chart administrations and 27.8% of electronic medication management administrations. Recorded times also tended to be earlier than observed administration times. For both paper charts and electronic records, the most common pattern was a documented time up to five minutes before the observed administration.
Larger timing discrepancies were more frequent with electronic records. With paper charts, the documented time was more than five minutes earlier than the observed administration time in 35.2% of administrations with a recorded time. With electronic medication management, this rose to 47.7%. Paper charts were more likely than electronic records to show a documented time more than five minutes later than observed administration, at 6.9% compared with 2.6%.
Electronic medication management supports more complete documentation in some areas and reduces administrations from incorrect, illegal or incomplete prescriptions. The same implementation also brings accuracy challenges, particularly when recorded administration details do not match observed practice or when documented times precede actual administration. Automated fields, changing workflows and conflicting electronic instructions may contribute to these discrepancies. Most administrations were documented correctly, but the higher rate of inaccurate details during implementation indicates a need for stronger training, workflow support and system configuration that allows nurses to record administration details accurately.
Source: BMJ Health & Care Informatics
Image Credit: iStock
References:
Badgery-Parker T, Li L, Woods A et al. (2026) Does the accuracy of medication administration documentation improve with electronic medication systems? A stepped-wedge cluster randomised trial. BMJ Health & Care Informatics, 33:e101507.
