Adverse drug events and medication errors in a paediatric inpatient population

Background. Medication-related patient injuries (adverse drug events, ADEs) are an important problem in all hospitalised populations; however, the potential for injury is reported to be greater in children than adults. Many ADEs are due to error and therefore could be prevented. Data regarding the risk factors (or predictors) for these events in paediatric inpatients is limited. It was hypothesised that quot;identification of risk factors for ADEs and medication errors in the paediatric inpatient setting will inform likely prevention strategiesquot;. Aims. To determine the frequency, nature and risk factors for ADEs and potential ADEs occurring in a paediatric inpatient population; to assess the vulnerable processes in the neonatal intensive care unit (NICU) medication use process; and to provide recommendations for the targeting of likely prevention strategies. Setting. A general paediatric ward (PW), postnatal ward (PNW) and NICU of a University- affiliated urban general hospital. Design. There were two study components: the medEVENT study which involved identification of actual ADEs and potential ADEs over a twelve week period, through prospective review of medical records, medication charts and administration records along with voluntary and solicited staff report and parent interview; and the FMEA study which used a proactive risk assessment technique, Failure Mode and Effect Analysis (FMEA), to rank all potential failures in the NICU medication use process according to risk. Results. In the MedEVENT study 3160 prescription episodes were reviewed (which represented 520 admissions, 3037 patient-days) and revealed a total of 67 ADEs and 77 potential ADEs. The greatest number of events occurred in NICU with very few events in the PNW. However, paediatric surgical admissions experienced the highest rate of ADEs per 1000 patient-days (80) as compared to medical (65) then NICU admissions (19). Over half of the ADEs were deemed preventable, 38 (57%), with the 'more serious' ADEs more likely to be preventable than 'not serious' ADEs. The impact on hospital resources was considerable with the cost attributed to extra bed days due to ADEs to be $NZD 50,000. Dosing errors were the most common type of error, particularly when prescribing and administering medications. Antibacterial and narcotic analgesics were commonly implicated, as was the intravenous route of administration. Few events were related to unlicensed use of medications. For ADEs, the major risk factors when analysed by admission, were greater medication exposure and increasing age; by prescription, were increasing age, oral route and narcotics and antibacterial agents; for paediatric ward admission, were increasing age and increased length of stay; and for NICU admission, no major risk factors emerged. For potential ADEs, the major risk factors when analysed by admission were greater medication exposure; by prescription, were junior prescriber, intravenous route, narcotics and antibacterials; for paediatric ward admission, were junior prescriber and narcotics; and for NICU admission were antibacterials, electrolytes and umbilical venous catheter administration. Neither ADEs nor potential ADEs were associated with unlicensed use of medicines or high alert status drugs. The FMEA study identified 72 potential failures in the NICU medication use process with 193 associated causes and effects. Multiple failures were possible in the process of 'prescribing medication' and in the process of 'preparation of medication for administration'. The highest ranking issues were found to occur at the administration stage. Common potential failures related to errors in the dose, timing of administration, infusion pump settings and route of administration. Conclusions. Analysis of the risk factors of ADEs and potential ADEs found that the most vulnerable processes were when prescribing and when preparing a medicine for administration; especially when involving narcotic and antibacterial agents and for children with greater medication exposure Strategies that selectively target these high risk areas are therefore likely to have the greatest impact on preventing drug-related injuries in hospitalised children