Medication errors in hospitals: computerized unit dose drug dispensing system versus ward stock distribution system.

Aim: The aim of this study was to evaluate the rates and types of drug prescription and administration errors in one pediatric nephrology ward, comparing two dispensing schemes : the first one defined as handwritten prescription plus ward stock distribution system (WSDS), and the second one as computerized prescription plus unit dose drug dispensing system (UDDDS). Method: Data were collected over an 8‐week period, from 1 February to 31 March 1999. Two fifth‐year pharmacy students photocopied prescription and administration documents on the ward each day, under the supervision of a senior pharmacist. The medical record analysis was used to compare the prescription with the administration report. Prescribing and administration medication errors were classified according to the American Society of Health‐System Pharmacists.Results: Prescribing errors: overall, for both dispensing schemes, a total of 511 prescriptions, resulting in 4532 prescribed drugs (an average of 9 drugs per prescription) were prescribed. The total prescription error rate was 20.7% (937 of 4532), resulting in 1.9 errors per patient per day. The computerized prescription error rate was 10.6% (419 of 3943), the handwritten prescription error rate was 87.9% (518 of 589). This difference was very significant (P < 0.0001).Administration errors: the total opportunity of administration errors was 4589 (sum of administered and omitted drugs). The total administration error rate was 23.5% (1077 of 4589) including wrong administration time, and 11.7% (538 of 4589) excluding administration time. The administration error rate, including administration associated with time errors, was only 22.5% (888 of 3943) for computerized prescription + UDDDS, compared with 29.3% (189 of 646) for handwritten prescriptions plus WSDS (P < 0.001). Excluding administration associated with time errors, the administration error rates were 9.7% and 24.3%, respectively (P < 0.0001). Conclusion: The drug prescription and administration error rates were significantly decreased using computerized prescription plus UDDDS as compared with handwritten prescription plus WSDS in a pediatric unit (even with potential biases taken into account).