5PSQ-147 Algorithms to enhance safety of high alert medications’ prescribing in hospitalised patients
Year: 2025
Type: conference-paper
Study design: Observational Study
Abstract: Background and Importance High alert medications can cause significant patient harm, when used inappropriately. To enhance patient safety, automated alerts are recommended by the Institute for Safe Medication Practices. Aim and Objectives To describe the use of automated alerts embedded in a computerised decision order system to identify drug related problems (DRP) for high alert medications in hospitalised patients, and associated pharmacists’ interventions. Material and Methods A retrospective study was conducted at two sites (academic and general) of a tertiary care hospital in Belgium. Seventeen algorithms were designed to identify DRP relevant for high alert medications and were implemented in our computerised decision order system. Alerts were based on the patients’ prescriptions and laboratory data. All alerts were checked by a pharmacist every day (except for weekends). All alerts occurring between September 2023 and August 2024 were analysed. Data collected included: alert type and whether the alert led to an actual DRP requiring a pharmacists’ intervention. Descriptive statistics were used to analyse the data. Results A total of 1147 alerts were analysed by the pharmacists, corresponding to a mean of 4.4 alerts daily. The positive predicted value was of 70.9%, as 813 alerts were actual DRPs and pharmacists’ interventions were required. Over half of the latter (53.3%, n=433) were related to antithrombotic agents: 23.7% (n=193) were related to overdosing antithrombotic (according to weight, renal function, age and INR), 27,6% (n=224) were related to underdosing (e.g. ‘once daily dosing’ of apixaban, or underdosing according to weight or INR) and 2,0% (n=16) were concomitant administrations of low molecular weight heparin and direct oral anticoagulants. A third of alerts (33,6%, n=260) were related to electrolytes, including potassium concentrations incompatible with peripheral intravenous administration, and potassium administration to patients without hypokalaemia. Insulin overdosing occurred for 5,5% (n=45) of alerts and 4,8% (n=39) of alerts resulted from administration of prolonged liberation opiates in patients with renal failure. Finally, 4,4% (n=36) of alerts were related to methotrexate, including multiple weekly administrations of the drug in inflammatory diseases for 3,2% (n=26) of alerts. Conclusion and Relevance Automated alerts tailored to identify DRPs for high alert medications, effectively identify opportunities for pharmacists’ interventions, enhancing patient safety. References and/or Acknowledgements Conflict of Interest No conflict of interest
Institution: CHU Dinant Godinne UCL Namur
Language: English
Subfield: Geriatrics and Gerontology
Field: Medicine
Domain: Health Sciences
Open Access status: gold
OpenAlex ID: W4408376436