This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.
Sudisth Kumar,
- Professor, Lecturer, Dept. Of Electronics and Communication Al Kabir Polytechnic Kopali,, Jamshedpur, India
Abstract
Medication-related errors remain a major challenge in healthcare, contributing to adverse drug events, increased hospitalization rates, and substantial healthcare costs. Conventional Clinical Decision Support Systems (CDSS) primarily rely on rule-based mechanisms for identifying drug-related problems (DRPs), including drug–drug interactions, contraindications, dosing errors, therapeutic duplication, and medication omissions. Although effective in structured environments, these systems frequently generate excessive context-insensitive alerts, leading to alert fatigue and reduced clinical acceptance. Recent developments in Artificial Intelligence (AI), Large Language Models (LLMs), and Retrieval-Augmented Generation (RAG) have created new opportunities for intelligent clinical reasoning and personalized decision support. However, standard RAG systems often suffer from performance degradation when exposed to noisy or partially relevant retrieved information. This study proposes RoseRAG-CDSS, an intelligent case-based clinical decision support framework that integrates retrieval-augmented language modeling, knowledge-guided reasoning, and margin-aware preference optimization for precision medication safety. The system functions as an AI-driven cognitive assistant capable of analyzing patient-specific clinical contexts, retrieving relevant medication knowledge, and generating explainable recommendations for healthcare professionals. To evaluate the framework, a dataset comprising 25 complex clinical case vignettes, 65 prescribing error scenarios, and multiple medical specialties was developed and validated by pharmacists and physicians. Drug-related problems were categorized using established clinical guidelines and assessed across multiple performance metrics including accuracy, precision, recall, F1-score, severity detection, and false-positive alert rates. Experimental findings demonstrate that RoseRAG-CDSS significantly improves robustness against retrieval noise while enhancing detection of clinically significant medication risks. The framework reduced false alerts, improved identification of severe drug-related problems, and supported more effective medication review when deployed in a pharmacist co-pilot setting. The proposed approach highlights the potential of integrating case-based reasoning, cognitive artificial intelligence, and knowledge-enhanced language models to develop trustworthy, explainable, and clinically actionable decision support systems for next-generation healthcare environments.
Keywords: Artificial Intelligence, Clinical Decision Support System, Retrieval-Augmented Generation, Cognitive Computing, Intelligent Agents, Drug-Related Problems, Medication Safety, Explainable AI, Healthcare Informatics
References
[1]. Anaya, Karim L., and Michael G. Pollitt. “Reactive power procurement: A review of current trends.” Applied Energy, vol. 270, issue C, no S0306261920304517, (2020). [2]. Esobinenwu, Chizindu Stanley and E. T. Oniyeburutan, “Reactive power (VAR) compensation techniques in high voltage transmission lines,” Global Journal of Engineering and Technology Advances, vol. 16, no. 01, (2023), pp. 024-029. [3]. Tabatabaei, N. M., A. Jafari and N. S. Boushehri, “A survey on reactive power optimization and voltage stability in power systems,” International Journal on Technical and Physical Problems of Engineering (IJTPE), vol. 6, no. 18, (2014), pp. 220-233. [4]. Ayalew, Frie, Seada Hussen and Gopi Krishna Pasam. “Reactive Power Compensation: A Review,” International Journal of Engineering Applied Sciences and Technology, vol. 3, no. 1, (2019), pp. 1-7. [5]. Ibekwe, B. E., C. A. Mgbachi and N. H. Nwobodo, “Reactive power compensation using 500kV, 180km line to maintain substantially flat voltage profile at all levels of AC transmission system,” International Journal of Advanced Research in IT and Engineering, vol. 6, no. 9, (2017), pp. 1-12. [6]. Zaidan, Majeed Rashid, and Saber Izadpanah Toos, “Optimal location of static var compensator to regulate voltage in power system,” IETE Journal of Research, vol. 69, no. 4, (2023), pp. 2177-2185. [7]. Bidyanath, Khaidem, Sanasam Dhanabanta Singh and Shuma Adhikari. “Implementation of genetic and particle swarm optimization algorithm for voltage profile improvement and loss reduction using capacitors in 132 kV Manipur transmission system.” Energy Reports, vol. 9, (2023), pp. 738-746. [8]. Bazilah Ismail, Noor Izzri Abdul Wahab, Mohammad Lutfi Othman, Mohd Amran Mohd Radzi, Kanendra Naidu Vijyakumar and Muhammad Najwan Mat Naai, “A Comprehensive Review on Optimal Location and Sizing of Reactive Power Compensation Using Hybrid-Based Approaches for Power Loss Reduction, Voltage Stability Improvement, Voltage Profile Enhancement and Loadability Enhancement”, IEEE Access Multidisciplinary Journal, (2020) vol 8, pp. 2169-3536 . [9]. Suliman, Mohammed Y, “Voltage profile Improvement using Static Synchronous Compensator STATCOM,” Journal of Information, Communication and Intelligence Systems, vol.1, no. 2, (2023), pp. 36-42. [10]. Aljendy, Raseel I., Rinat R. Nasyrov, Hamdy M. Sultan and Ahmad A. Zaki Diab. “Bio-inspired Optimization Techniques for Compensation Reactive Power and Voltage Profile Improvement in the Syrian Electrical Distribution Systems.” Proceedings of the IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering (EIConRus), pp. 928-933. IEEE, 2019https://doi.org/10.1109/EIConRus.2019.8656900 28-31 January 2019 IEEE Conference (2019) Conference Location: Saint Petersburg and Moscow, Russia.
Journal of Artificial Intelligence Research & Advances
| Volume | 13 | |
| 03 | ||
| Received | 18/04/2026 | |
| Accepted | 04/08/2026 | |
| Published | 30/09/2026 | |
| Publication Time | 165 Days |