Background and Aim Medication errors represent the largest proportion of all types of medical errors.1 They occur at all stages of patient management.2 Limited evidence and popular belief suggests that they are more common in neonatal and paediatric patients than in adults. A 2006 systematic review of studies identifying the rate of medication errors in paediatric patients found only 32 relevant articles.3 We are aware that the literature on this subject has increased dramatically in the last few years. The aim of this study was to identify and explore all published studies since 2006 which determined the rate and/or number of medication errors in neonates and children.
Methods Six databases were searched (BNI, EMBASE, IPA, MEDLINE, AMED and CINAHL) from April 2006 to March 2011.
Inclusion criteria: original research studies identifying the rate or number of medication errors in neonatal or paediatric settings or in the general population where neonatal or paediatric data were separately identified.
Results 153 papers met the inclusion criteria. The studies involved:
29 countries—mostly the US (83 studies) followed by the UK (16 studies)
22 different settings—mostly specialist children's hospitals followed respectively by neonatal units and paediatric units in general hospitals.
68 different interventional tools to try to reduce the rate of error.
234 different definitions of medication errors in 76 studies (the remaining 77 studies did not use a definition).
25 different denominators to express the data collected. Commonest denominators were “of all errors”, “of all orders” and “of all patients”. Many studies did not use a denominator and only identified the number of errors.
5 different methods to collect data (chart/medical record review (39%), review of incident reports (32.5%), mixed methods (14.3%), simulation (8.4%) and direct observation (5.8%)).
identification of different types of errors: prescribing errors (29.2%), administration errors (23.4%), dispensing errors (0.65%), monitoring errors (0.65%), medication errors generally (18.8%) and a mixture of different types of medication errors (27.3%).
all medications (73 studies) and specific medications only (80 studies).
Conclusion This review shows that the literature on the subject of neonatal and paediatric medication errors has increased dramatically since 2006. The published studies are however highly heterogeneous making it very difficult to compare studies and to provide an accurate estimate of neonatal or paediatric medication errors in a particular setting. The use of different denominators to express the data made comparison particularly difficult. Further work is needed to explain the influence of the above factors on documented error rates.