Characteristics of adverse medication events in a children's hospital

Aim To compare adverse medication events ( AME s) reported in children, via the I nternational S tatistical C lassification of D iseases and R elated H ealth P roblems 10th Revision ( ICD ‐10) coding with events reported via other data sources. Method AME reports were retrieved using codes Y 40– Y 59 and X 40– X 44 over 6 months. Patients’ charts were manually reviewed to identify events associated with error and/or harm with medicines during a hospital admission. Medication name, group, error, harm and alert documentation were recorded. Clinical incidents and pharmacist interventions were reviewed for the same period. Results Two hundred sixty‐three events from J anuary to J une 2011 were recorded by ICD ‐10 coding in 180 patients. After duplicated, missing or unrelated events were excluded and 146 AME s remained. In the same period, 117 AME s were reported as incidents and 190 as pharmacist interventions. In total, 276 children with 447 events were reported via all sources. Little duplication between data sources was evident. In total, 158 events involved harm, with 135 of these from ICD ‐10 coding, 16 from incident reports and 2 pharmacist interventions (including 6 events from multiple sources). Error was involved in 3% of ICD 10 reports, 97% of incidents and 100% of interventions. Only 14% of harm‐related events from ICD ‐10 were documented on the medical record clinical alert. Chemotherapy accounted for 31% of harm‐related events, antimicrobials 18%, corticosteroids 14% and narcotics 12%. Conclusion Of the harm‐related events, 85% were documented via ICD ‐10 coding with few documented in other databases. Review of ICD ‐10‐coded AME s can provide valuable information to improve patient safety and quality.