传染病与多重耐药菌感染智能识别防控系统的构建及应用

Construction and application of an intelligent system for recognition, prevention, and control of infectious diseases and multidrug-resistant orga-nism infections

摘要:

目的 以提升全流程管理成效为导向构建传染病与多重耐药菌感染的智能识别防控系统,研究其应用效果。方法 基于实时精准反映感染状态与传播风险的个性化逻辑解析规则,搭建具有自动识别、动态标记、实时共享、前置预警及可视化指引功能的智能识别防控系统。选取2023年10月—2024年5月某三级甲等医院开展侵入性诊疗操作的两个平台科室的就诊患者为研究对象,采用自身对照法,按照传统人工模式管理为对照组,应用智能系统管理为试验组,比较系统应用前后识别防控效果的差异。结果 共纳入研究对象2 146例患者,应用智能系统后的感染者识别防控率和感染者识别准确率均较人工模式明显提升(分别从5.3%、72.4%提高至100%),差异均有统计学意义(均P＜0.001)。感染信息中位前置预警时间达85.20 d,且实现100%前置预警。医护人员在感染信息识别管理环节的平均耗时节省4.71 h/d。结论 本研究搭建的智能系统可显著改善传染病与多重耐药菌感染者从识别到防控环节的全流程管理成效,有效降低交叉感染风险,提高诊疗服务效率。

Abstract:

Objective To construct an intelligent recognition, prevention, and control system for infectious diseases and multidrug-resistant organism infections, aiming at improving the efficacy of full-process management, and to evaluate its application effects. Methods Based on personalized logic parsing rules that accurately reflect the infection status and transmission risks in real-time, an intelligent recognition, prevention, and control system with functions of automatic recognition, dynamic labeling, real-time sharing, early warning, and visual guidance was established. Patients undergoing invasive diagnostic and therapeutic procedures in two departments of a tertiary first-class hospital from October 2023 to May 2024 were selected as the research subjects. The differences in recognition, prevention, and control efficacy before and after the application of the system were compared using a self-controlled method, with traditional manual management as the control group and intelligent system management as the experimental group. Results A total of 2 146 patients were included in the analysis. The recognition, prevention, and control rate and the accuracy rate of recognizing infected individuals using the intelligent system were enhanced significantly compared with those using manual mode (improved from 5.3% and 72.4% to 100%, respectively), with statistical significance (both P<0.001). The median early warning time for infection information reached 85.20 days, with 100% early warning achieved. The average time spent by medical staff on infection information recognition and management was reduced by 4.71 hours per day. Conclusion The intelligent system constructed in this study significantly improves the effectiveness of full-process management in recognition, prevention, and control of infectious diseases and multidrug-resistant organism infection, effectively reduces the risk of cross-infection, and enhances the efficiency of diagnostic and therapeutic services.