Background The ward drug trolley process is error-prone so therefore pharmacists should take measure to recognise and prevent them.

Purpose To evaluate the rates and types of dispensing errors (DE) during the drug trolley process.

Materials and methods Prospective observational study. Data were collected for 44 working days in 2009 and 2010. The hospital had 350 beds with seven medical and four surgical wards. 84.3% of beds use a unit-dose dispensing system (UDDS) plus written transcription (UDDS-WT) and 15.7% use UDDS plus computerised prescription order entry (UDDS-CPOE). Each day pharmacists randomly selected one or two trolleys and checked them. Dispensing errors were classified as: Type 1: wrong patient, Type 2: omission of drug, Type 3: drug not prescribed, Type 4: Wrong dose, route or dosage form and Type 5: Quantity error. The authors calculated the dispensing error rate (DER) by dividing DE by the opportunities for error (OE: total units dispensed+doses prepared in the drug trolleys).

Results The observations were conducted on 56 drug trolleys (1928 beds): 36 medical and 20 surgical, 14 428 total of doses prepared and dispensed (OE). 137 DEs were detected: 1.46% (2) type 1, 40.15% (55) type 2, 32.17% (44) type 3, 13.14% (18) type 4 and 13.14% (18) type 5. The most frequent errors are type 2 and type 3, related to the prescription changes after the drug trolley process. The DE rate was 0.95% (137 of 14,428). The DE rate in UDDS-WT was 0.91% (117 of 12868) and in UDDS-CPOE it was 1.28% (20 of 1560).

Conclusions The short period of our study and the great difference in the methodology used in other studies hinder the comparison with their results. Although there are almost no differences between our DE rate in the two modalities of dispensing, it was not possible to compare them as the OE varied substantially. Despite the fact that the DE rate was low, recognising the incidence and types of medication errors allows us to analyse the causes to help achieve maximum patient safety.