سامانه همایش های دانشگاه آزاد اسلامی واحد نجف آباد

صفحه اصلی / اولین همایش بین المللی تحول خدمات سلامت: تلفیق مراقبت، درمان، بازتوانی و امیدبخشی

The Role of Artificial Intelligence in Preventing Medication Errors and Enhancing Medication Safety: Opportunities, Challenges, and Future Perspectives

نویسندگان :

Hamed Moeini¹ Mina Allami² Mina Jouzi³

1- دانشگاه آزاد اسلامی واحد نجف آباد 2- دانشگاه آزاد اسلامی واحد نجف آباد 3- دانشگاه آزاد اسلامی واحد نجف آباد

کلمات کلیدی :

Artificial Intelligence،Medication Errors،Machine Learning،Patient Safety،Drug-Related Side Effects and Adverse Reactions،Decision Support Systems،Clinical

چکیده :

Background : This review critically analyzes the part of artificial intelligence (AI) in medication errors prevention and improvement of medication safety. It assesses the feasibility of using machine learning, natural language processing, and newer AI technologies for adverse drug event prediction and detection, medication-error identification, and clinical decision support . (1, 2) It also looks at the use of AI with electronic health records and the disconnect between the performance of algorithms and observed clinical effectiveness . (1, 3) Lastly, the review highlights clinical, technical, and implementation challenges of safe implementation, generalizability, and clinical use of AI-based medication-safety systems and collates potential new opportunities for the responsible use of AI in the clinical context . (1, 3) Method : A critical narrative review was used to draw together the evidence for the use of AI applications for medication-error prevention and medication safety. AI and machine learning, natural language processing, generative AI, adverse drug events, clinical decision support and pharmacovigilance were critically evaluated according to application, safety outcome, clinical relevance, and reported limitations of the relevant studies. Findings : AI has potential applications in medication safety, such as ADE prediction and identification, medication error identification, clinical decision support, and pharmacovigilance, especially in the context of EHR data.(1) Though there is promise shown in ADE prediction performance, the standardization and clinical evaluation in real-world settings still need to be addressed for ML . (2) NLP can assist with supporting the detection of ADEs from unstructured clinical text (3) and emerging LLM applications can be used to assist with ADE detection and decision support, but there remains a need to assess the reliability and clinical applicability of these applications . (4) AI has also identified actual medication mistakes in the real world, such as vial-swap mistakes before giving meds . (5) But because of heterogeneity, there is limited external validation and the evidence is mostly retrospective limiting generalizability and clinical translation . (1) Therefore, with a view to broader uptake, there is a need for strict validation, transparency and field testing . (1, 2, 4) Conclusion : AI tools have great potential to enhance medication safety by predicting and detecting ADEs, identify medication errors, and provide clinical decision support. However, there are some restrictions in the validation, generalizability and implementation in clinical practice to date that limit the widespread use of this. Research should focus on future prospective and external validation of the systems, reporting and ensuring transparent communication, and effective integration into clinical routine to maximise the value of AI-based systems for medication safety in real-world settings . (1, 2, 4)