Prevalence and determinants of medication administration errors in clinical wards: A two-centre prospective observational study.

This was an observational study, so it can show associations but not prove that a determinant causes an error. Wrong time errors and procedural errors were excluded, so the study did not capture all medication administration problems. The study measured potential harm rather than actual harm to patients. Staff characteristic data were unavailable for about 45% of observations, so associations with nursing staff characteristics are uncertain. Observer and reviewer bias remained possible despite disguised observation and training. The study was done in two hospitals before later medication safety changes, and it did not examine all possible workload, patient or environmental factors.

Do not conclude that electronic medication systems make medication administration safe. Errors occurred in 13.7% of administrations in hospitals using an electronic medical record, computerised physician order entry and an electronic medication administration record.