Optimization of electronic prescribing in pediatric patients

Improving pediatric patient safety by preventing medication errors that may result in adverse drug events and consequent healthcare expenditure,is a worldwide challenge to healthcare. In pediatrics, reported medication error rates in general, and prescribing error rates in particular, vary between studies, depending on the definitions and study methods used, and the setting studied. This thesis focuses on prescribing errors in pediatric patients and the use of electronic prescribing systems to prevent them. The nature, frequency and determinants of medication prescribing errors in pediatric inpatients were determined and the effect of computerized physician order entry (CPOE) and clinical decision support (CDS) on these errors were studied in the Wilhelmina Children’s Hospital, Utrecht. Extra attention was paid to pediatric and neonatal intensive care unit (PICU and NICU) patients as these offer extra challenges due to their several complex health problems and multi-drug treatments. Approximately 1% of electronic medication orders in the children’s hospital had to be intervened by the clinical pharmacy due to an error: about 80% concerned a correction and about 20% a completion. The majority of the corrections concerned a wrong dose or a wrong drug formulation. The majority of the completions concerned absent body weight, dosage form or strength/concentration of the prescribed drug. Free-text entry, the youngest of age and the oral dosage form and -route of administration were associated with prescribing errors. Focusing on the PICU, both handwritten and electronic medication orders were examined: 18% contained administrative errors, 53% omissions and 12% dosing errors. Additionally, in almost 20% of patients at least one potential drug-drug interaction (pDDI) was identified during PICU admission. Writing by hand, alterations in existing medication orders, intermittent dosing and ‘on demand use’ were identified as most important risk factors for prescribing errors. The effects of CPOE /CDS systems on medication prescribing errors, ADEs, and mortality in inpatient pediatric care and neonatal and pediatric intensive care settings were reviewed. Overall, CPOE/CDS systems clearly reduced medication prescribingerrors, if well-designed and -implemented. However, effect on clinically relevant outcomes could notbe demonstrated, possibly due to a limited set of outcome data restricted to pediatric and neonatal data. In an attempt to contribute to the evidence base, the effects of advanced CPOE/CDS for glucose control in NICU patients were studied, focusing on hypo- and hyperglycemic episodes (a clinically relevant outcome for this population) and prescribing time efficiency. The studied computerized prescribing and calculating CDS tool proved to preserve accuracy for calculation and control of glucose intake and decrease time needed to prescribe. To optimally prevent prescribing errors using CPOE/CDS in children, the systems need to be more advanced and better tailored to pediatric care, preferably based on clinical experience and scientific evidence. System requirements and design of an electronic prescribing system for PICU and NICU is presented, including testing of the underlying model. The developed system aims to be integrated, safe by default and efficient and has the potential to solve several of the main problems related to the medication process in such specific patients.