2025 Southeastern Residency Conference

Title: Evaluation of Pediatric Medication Safety Mechanisms Within a Newly Implemented Electronic Health Record in the Emergency Department of an Adult Community Hospital


Authors: Anne Thomas Hooper, Anna Beth Bowles, Kevin Vanlandingham, Elizabeth Davidson


Background: Medication errors are a predominant factor in preventable patient harm. Pediatric patients are especially vulnerable due to unique medication dosing and administration principles. Electronic Health Record (EHR) systems, designed primarily for adult populations, can fail to detect errors in pediatric medication orders. The fast-paced environment, use of verbal orders, and critical nature of patients are factors that further increase risk of medication errors in the emergency department. This study aims to assess the efficacy of EHR safety mechanisms in identifying pediatric medication errors using a computerized provider order entry (CPOE) system in an adult hospital emergency setting.


Methods: The Plan-Do-Study-Act (PDSA) methodology was used to evaluate and adapt EHR safety mechanisms for quality improvement of pediatric medication safety in CPOE. A literature review was used to identify gaps in pediatric medication safety of the EHR and CPOE within high risk areas of hospital organizations. Resources such as the KIDs List, National Pediatric Readiness Program, and the Leapfrog evaluation tool provided the framework for the development of test questions. The "Do" phase involved ordering medications on test patients that were expected to trigger a safety alert. Questions to assess current medication safety alerts were based on three categories: weight-based dosing, maximum recommended dosing, and miscellaneous. Miscellaneous questions included incorrect routes, forms, and excipients. The “Study” portion involved recording results and analysis of the final endpoints. The primary endpoint is the frequency of medication order entry errors. The secondary endpoint was the detection frequency of medication errors within the categories. A “pass” was recorded if an appropriate alert populated and a fail was recorded if an alert was undetected based on the predetermined questions. During the “Act” phase, findings were assessed and changes implemented to the EHR where feasible. 


Results: In Progress


Conclusion: In Progress