P31 The effect of electronic prescribing on the incidence and nature of paediatric medication errors – an observational study

Abstract

BackgroundPrescribing errors may adversely affect the safety of patients of all ages and particularly paediatric patients due to the complexity of calculating individual patient doses and the common need to use medicines not designed for children.Electronic prescribing systems (EPS) are an intervention used to try reduce such errors. Few studies have evaluated their impact on prescribing errors in the UK particularly in neonatal/paediatric settings.AimsTo evaluate the nature and prevalence of prescribing errors in a UK children’s hospital which uses an EPS.To compare the results with those of a previous study1 conducted in the same hospital before the introduction of EPS.MethodsPharmacists were shadowed in their routine clinical work by the researcher who documented all prescribing errors identified. All electronic prescriptions checked for patients≤18 years in the paediatric wards were included in the data collection. Data was entered into a SPSS database for analysis.Types and severity of errors identified were categorised according to criteria used in a previous General Medical Council funded study (EQUIP).2 Results were compared to a previous study from 2012/2013 in the same hospital (4204 prescriptions) using the same method before the introduction of EPS.ResultsObservation of 4035 prescriptions checked by pharmacists on their routine ward visits (754 patients) during September-December 2015 identified 208 prescribing errors, an error rate of 5.2% prescriptions.There was a statistically significant difference between the rate of errors identified in this study compared to that found in the written prescriptions (8.7%) in the previous study (p<0.001).Fifteen errors (7.2%) were classified as serious; 106 errors (51%) significant and minor errors represented the remaining errors (data unavailable for previous study).The most common types of error identified were omission of regular medications (1.7% total prescriptions vs 1.6% in previous study); under-doses (0.8% vs 0.7%), non-measurable doses (0.7% vs 1.5%) and overdoses (0.5% vs 0.7%). Illegible prescriptions were reduced from 2% total prescriptions in the first study to zero by the EPS. Errors involving incorrect/missing administration times were also reduced from 0.9% in the previous study to 0.2% in this study.Antibacterial, analgesics and bronchodilators were the most common groups of drugs associated with errors (28.8%, 10.6% and 10.6% of total errors respectively). The oral route was most commonly involved in errors (47.2% of total errors) followed by intravenous (30.3%) (data unavailable for previous study).ConclusionThe EPS was associated with a significant reduction in prescribing errors compared with written prescriptions. Illegible prescription errors were eliminated and incorrect/missing administration time errors greatly reduced. Errors involving non-measurable doses were more than halved. EPS however had a negligible effect on under- and over-dosing errors. Introduction of a more sophisticated clinical decision support system may help to reduce ...