MTP4.08 Advances and Future Directions in Burn Fluid Resuscitation: A Comprehensive Literature Review

Background Optimal fluid resuscitation in burn patients is crucial to prevent fluid overload while ensuring adequate euvolemia. Traditional methods, such as the Parkland and Modified Brooke formulas, primarily utilize urine output (UO) to guide therapy but lack patient-specific customization and are inherently reactive. With advancements in artificial intelligence (AI), there is potential for more tailored and proactive fluid management strategies. Methods This review assesses the incorporation of AI in burn fluid resuscitation, emphasizing both historical rule-based systems and emerging machine learning (ML) models. Literature searches were conducted following JBI guidelines and the PRISMA-ScR checklist, focusing on studies that employ AI and automated technologies in managing burn fluid therapy. Results Among 149 identified studies, 22 met the inclusion criteria. These studies highlight the limited application of ML in real-time management, with existing AI models primarily relying on rule-based algorithms like the Burn Navigator® (BN). Innovations in simulation algorithms and real-time monitoring devices provide a groundwork for AI integration, facilitating a transition from traditional formulas to AI-driven models. Conclusion The evolution from static formulas to dynamic algorithms has marked significant progress in burn resuscitation. While traditional resuscitation formulas provide a foundational approach, AI integration promises personalized and adaptive fluid management. Future directions involve enhancing AI applications within clinical settings to improve outcomes in burn care by reducing complications associated with fluid overload and adapting to patient-specific needs.