Research Center for Healthcare Data Science, Zhejiang Laboratory, Hangzhou, China.
Emergency Department, the Second Affiliated Hospital, Zhejiang University School of Medicine and Institute of Emergency Medicine, Zhejiang University, Hangzhou, China.
Comput Methods Programs Biomed. 2023 Jun;235:107512. doi: 10.1016/j.cmpb.2023.107512. Epub 2023 Mar 26.
For severe trauma patients, hemorrhage is the most common cause of medically preventable deaths. Early transfusion is beneficial to major hemorrhagic patients. However, the early supply of emergency blood products for major hemorrhagic patients is still a major problem in many areas. The aim of this study was to design and develop an unmanned emergency blood dispatch system for the fast delivery of blood resources and rapid emergency response to trauma events, especially those with mass hemorrhagic trauma patients and those occurred in remote areas.
Based on the process of emergency medical services for trauma patients, we introduced unmanned aerial vehicle (UAV) and designed the main flowchart of the dispatch system, which combines an emergency transfusion prediction model and UAV-related dispatch algorithms to improve first aid efficiency and quality. The system identifies patients in need of emergency transfusion through a multidimensional prediction model. Then, by analyzing the blood center, hospitals and UAV stations nearby, the system recommends the patient's transfer destination for emergency transfusion and dispatch schemes of UAVs and trucks for a fast supply of blood products. Simulation experiments of urban and rural scenarios were conducted to evaluate the proposed system.
The developed emergency transfusion prediction model of the proposed system achieves a higher AUROC value of 0.8453 than a classical transfusion prediction score. In the urban experiment, by adopting the proposed system, the average wait time per patient decreased from 32 to 18 min, and the total time decreased from 42 to 29 min. Owing to the combination of the prediction and the fast delivery function, the proposed system took 4 and 11 min less wait time than the strategy with only the prediction function and the strategy with only the fast delivery function, respectively. In the rural experiment, for trauma patients requiring an emergency transfusion at 4 locations, the wait time for transfusion under the proposed system was 16.54, 17.08, 38.70 and 46.00 min less than that under the conventional strategy. The health status-related score increased by 6.9%, 0.9%, 19.1% and 36.7%, respectively.
Experimental results demonstrate that the proposed system works well with a faster blood supply speed for severe hemorrhagic patients and better health status. With the assistance of the system, emergency doctors at the scene of an injury are able to comprehensively analyze patients' status and the surrounding rescue conditions and then make decisions, especially when encountering mass casualties or casualties in remote areas.
对于严重创伤患者，出血是最常见的可预防的医源性死亡原因。早期输血对大出血患者有益。然而，为大出血患者提供紧急血液产品的早期供应仍然是许多地区的一个主要问题。本研究旨在设计和开发一种无人应急血液配送系统，以便快速运送血液资源，并对创伤事件做出快速应急反应，特别是针对大量出血性创伤患者和发生在偏远地区的创伤事件。
基于创伤患者的紧急医疗服务流程，我们引入了无人机（UAV）并设计了调度系统的主要流程图，该系统结合了紧急输血预测模型和与无人机相关的调度算法，以提高急救效率和质量。该系统通过多维预测模型识别需要紧急输血的患者。然后，通过分析附近的血液中心、医院和无人机站，系统为患者推荐紧急输血的转院目的地，并为无人机和卡车的快速供应血液产品制定调度方案。针对城乡场景进行了仿真实验，以评估所提出的系统。
所开发的系统的紧急输血预测模型的 AUROC 值达到了 0.8453，高于经典的输血预测评分。在城市实验中，通过采用所提出的系统，每位患者的平均等待时间从 32 分钟减少到 18 分钟，总时间从 42 分钟减少到 29 分钟。由于预测和快速配送功能的结合，与仅采用预测功能的策略和仅采用快速配送功能的策略相比，所提出的系统分别减少了 4 分钟和 11 分钟的等待时间。在农村实验中，对于 4 个位置需要紧急输血的创伤患者，在所提出的系统下的输血等待时间比常规策略分别减少了 16.54、17.08、38.70 和 46.00 分钟。健康状况相关评分分别提高了 6.9%、0.9%、19.1%和 36.7%。
实验结果表明，所提出的系统对于严重出血患者的血液供应速度更快，患者的健康状况更好。在该系统的帮助下，创伤现场的急救医生能够全面分析患者的状况和周围的救援条件，然后做出决策，特别是在遇到大量伤亡或发生在偏远地区的伤亡时。
