Abstract
INTRODUCTION: Medication errors continue to cause inpatient harm in children and can be difficult to both identify and classify. Medication error studies often focus on assessing potential harm and there is little published data on actual harm from medication errors in children. OBJECTIVE: Our aim was to use multidisciplinary panels to identify and describe the actual harm resulting from prescribing and administration medication errors occurring at a major paediatric hospital. METHODS: We reviewed medication error data collected from retrospective medication record reviews to identify prescribing errors (26,369 orders, 19,692 errors and 3782 patients) and prospective direct observations (5137 dose administrations, 3663 errors and 1530 patients) to identify administration errors. Errors with the potential to cause serious harm and with evidence that the error reached the patient formed the dataset for our study. Case studies (n = 566) describing the prescribing and administration errors and a brief clinical summary were reviewed by multidisciplinary panels to determine whether there was evidence in patients' records of actual harm and to rate the severity of the harm identified. RESULTS: Actual harm was identified in 89 case studies and rated as minor in 43% (n = 38), moderate in 48% (n = 43) and serious in 9% (n = 8). There were no cases of harm rated as severe resulting in death. Antibacterials were the most common medications in cases with harm (n = 38/89 cases), and dosing errors (n = 32/89) the most common error type associated with harm. Younger patients had a significantly (t = 2.4, df = 198, p = 0.017) greater risk of actual harm from medication errors, and children aged under 12 months formed a higher proportion of those with actual harm (χ(2) (1, N = 566) = 10.5, p = 0.001). The most frequent type of administration errors leading to harm were wrong infusion rates of intravenous antibiotics (19/67 cases); 12 of these instances occurred in children under 12 months. Administration errors were more likely to result in actual harm (1.83%; 67 /3663 errors) compared with prescribing errors (0.21%; 42/19,692). CONCLUSIONS: We found higher rates of actual harm associated with medication errors in younger patients, wrong dose prescribing errors and intravenous antibiotic administration errors. These important findings provide opportunities for developing tailored interventions targeting identified high-risk areas to enable the successful reduction of preventable harms in paediatric patients.