Background Medication errors result in significant patient harm and are estimated to cost the NHS £750 m per year.1 Recent studies have shown that an excess of these errors occur in Paediatric Intensive Care Units (PICU).2 Error rates of 11.1% in the USA3 and 13.2% in the UK4 have been reported. In our PICU baseline rates were between 12% and 26% with evidence of a seasonal variation associated with higher workload. In addition infusions were noted to make up a significant proportion of these errors. A number of interventions have been introduced in response to the findings of previous audits namely the indroduction of specialised prescribing areas and preprinted forms for continuous IV infusions. This study was undertaken to evaluate the impact of these interventions.
Method Retrospective chart review of all active prescriptions over a 4 week period. Once only medications and IV fluids were excluded. Errors were classified by type and potential severity and results were compared with the previous data.
Results 1000 prescriptions were reviewed, showing an overall error rate of 14%, a reduction from the 26% found with a similar workload. 80% of errors had low or no potential adverse effects, as compared to the 90–95% found previously. Of the remaining errors, 13 were deemed potentially significant, and 7 potentially serious. This is higher than the previous reports. Errors in continuous IV infusions dropped from 36% to 23.4%.
Discussion Error rates based on occupied bed days are comparably lower than periods with similar workloadssuggesting a positive benefit following the interventions. Fewer transcription errors show the positive effect of zero tolerance policies. Pre-printed orders for commonly used medications have helped reduce infusion errors. 87% of errors, however, were related to a failure in communication; this provides scope for further work to improve the delivery of written orders. We believe that the nature of the more serious errors, seen in increased numbers during this study, might be addressed through the provision additional decision support infrastructure which will be the subject of ongoing work.