Introduction: Adverse drug effects are a significant public health problem. Prescription errors are responsible for a significant proportion of these adverse effects.
Methods: We have aimed to improve the link between generation of and interpretation of a prescription through computerisation. The prescription sheet, which is generated daily, was organised to allow care planning without the need to re-copy out treatments on the sheet. A prescription aid was available which was based on a core group of drugs commonly used in our respiratory service. The aim of the study was to compare the various types of errors observed during 6 weeks of computerized prescriptions (229 files) to a retrospective series of handwritten prescriptions of the service at an identical time (184 files) the previous year. The case-mix was identical for both analysed periods.
Results: The total number of technical prescribing errors in the 1,599 handwritten lines (49.27% error) was significantly higher (p<0.001) than the 1,805 computerized prescriptions lines (42.88% error). The errors of copying (p<0.001), eligibility (p<0.001) and incorrect spelling (p<0.05) were the main sources of error which were significantly reduced by computerisation.
Conclusion: Computerised prescription is likely to reduce the incidence of prescribing errors and adverse drug effects.