5PSQ-123 Reported serious adverse drug events: a knowledge source for pharmaceutical decision support systems learning

<h3>Background and Importance</h3> Drug iatrogenia is responsible for 10,000 deaths per year in France. The reporting of Serious adverse drug events (SADE) lead to retrospective implementation of barrier measures aimed at reducing their occurrence. The analysis of SADEs and the transcription of risk situations into a pharmaceutical decision support system (PDSS) would contribute to securing patient medication care in an innovative way. <h3>Aim and Objectives</h3> To present the importance of using databases of SADE as a source of pharmaceutical algorithms (PAs) utilised by PDSS. <h3>Material and Methods</h3> Two databases of SADEs were analysed by three clinical pharmacists: a key elements synthesis relating to prescription errors identified in the National Database of reporting serious adverse events related to care provided by the French National Authority for Health and a sample of SADEs notified to the Mutual insurance of the French health corps. The potential for formalising each SADE into PAs was evaluated based on several aspects: the existence of a drug related problem (DRP) with possible solutions, the explicit, encodable, and queryable nature of the assessment elements to represent the SADE in a PA and the availability of related health data in the PDSS. <h3>Results</h3> With a sample of 125 SADEs, 57 events (45.6%) were identified as potentially detectable and preventable (53) or mitigable through the creation of 34 PAs. Most of them involved cardiology drugs (8 PAs, 23.5%), followed by endocrinology (5) and analgesics (6). The primary DRPs identified were contraindications based on patient history (9 PAs, 26.5%), such as allergies and organ deficiencies; untreated indications (5) (e.g., absence of anticoagulants, antiplatelet agents); overdoses (5) (including anticoagulants and acetaminophen); and inadequate monitoring (liver enzymes, electrolyte panels). <h3>Conclusion and Relevance</h3> SADE reporting system seems a valuable source of PAs for spreading detection of these situations causing patient harm. The underreporting of these events related to prescription, monitoring, or side effects remains a significant issue that could be addressed through the educational and innovative applications of artificial intelligence. Nevertheless, additional solutions are still necessary to reduce SADEs resulting from drug administration processes. <h3>References and/or Acknowledgements</h3> Auger C, <i>et al</i>. Serious adverse events in France: a reporting and learning system. <i>International Forum on Quality &amp; Safety in Healthcare</i>. Copenhagen 2020 <h3>Conflict of Interest</h3> No conflict of interest