Objective: To estimate the proportion of medication errors in a tertiary hospital, global and for each delivery medication system, to describe the error types and the implied medications, and to analyze the factors associated to the same ones.
Methods: Errors were identified from direct observation of 2,242 opportunities for error (administered doses or prescribed doses not given) by 6 couples of observers. Delivery medication systems were stock in ward, unit dose with electronic prescription and unit dose with computerized transcription. Logistic regression was used to evaluate the association between errors and certain factors.
Results: The medication error rate was of 7.2% (CI 95%: 6.1-8.3), and 4.4% (CI 95%: 3.6-5.3) of them reached the patient. For delivery systems, the error rate was of 9.5% (CI 95%: 7.4-11.9) for stock in ward, 7.8% (CI 95%: 5.9-10.0) for electronic prescription and 4.7% (CI 95%: 3.4-6.4) for computerized transcription. The highest error frequency was observed in the administration phase (58.4%) and the omitted dose was the most prevalent error (31.7%). The error rate was associated to the pharmacotherapeutic process, the schedule of administration and the unit of hospitalization.
Conclusions: In one of each 14 opportunities for error a medication error takes place. The different delivery medication systems have different error rates.