医院内骨质疏松性骨折信息“自动抓取数据库”构建及12754例患者智能化管理应用报告  

Development of an automated osteoporotic fracture data retrieval database and its application in intelligent management of 12754 patients:an in-hospital report

作　　者：朱柯雨 刘功稳 王一可 魏祺 郑苗 翁程伟 云思敏 王雄毅 张东[1] 徐又佳[1,2] ZHU Ke-yu;LIU Gong-wen;WANG Yi-ke;WEI Qi;ZHENG Miao;WENG Cheng-wei;YUN Si-min;WANG Xiong-yi;ZHANG Dong;XU You-jia(Department of Osteoporosis Clinical Center,Second Affiliated Hospital of Soochow University,Suzhou 215000,Jiangsu,China;Department of Orthopedics,Second Affiliated Hospital of Soochow University,Suzhou 215000,Jiangsu,China;Department of Orthopedics,Suzhou Hospital of Traditional Chinese Medicine,Suzhou 215009,Jiangsu,China)

机构地区：[1]苏州大学附属第二医院骨质疏松症临床中心,江苏苏州215000 [2]苏州大学附属第二医院骨科,江苏苏州215000 [3]苏州市中医医院骨科,江苏苏州215009

出　　处：《中华骨质疏松和骨矿盐疾病杂志》2025年第1期60-72,共13页Chinese Journal Of Osteoporosis And Bone Mineral Research

基　　金：国家自然科学基金(82372455);江苏省医学重点实验室建设单位(JSDW202254);“中核医疗‘核医科技创新’计划资助项目”(ZHYLZD2023001);苏州大学研究生教育改革成果奖培育项目;江苏省体医融合促进老年骨骼健康应用工程研究中心。

摘　　要：目的为提高医院内骨质疏松性骨折(osteoporotic fracture,OF)患者诊疗质量和管理效率,本研究自主构建一种医院内自动抓取相关资料的“骨质疏松性骨折数据库”,数据库内置管理流程相关的智能化功能模块。在此基础上,分析该数据库在实际场景应用的结果和有效性。方法构建院内封闭式多源异构数据整合的专病数据库,数据库接口可后台对接医院的信息系统(hospital information system,HIS)、影像归档和通信系统(picture archiving and communication systems,PACS)、实验室信息系统(laboratory information system,LIS)等固有数据平台,并自动运用自然语言处理(natural language processing,NLP)技术识别及整合OF患者相关信息。运用该数据库纳入2022年6月至2024年6月苏州大学附属第二医院收治的50岁以上、4部位骨折(椎体、髋部、肱骨近端和桡骨远端)的12754例患者,并对患者信息进行智能化管理应用分析。结果该数据库可按照纳入条件自动获得12754例患者数据,并自动收集患者基本资料、病历或影像检查的骨折记录、检验检查结果、实时治疗方案等407个结构化字段信息。数据库可自动完成患者的骨质疏松相关数据识别(骨折部位、骨密度值、骨代谢相关指标、抗骨质疏松药使用)、院内转科及经治医生追踪、院内多次骨折记录检索。当患者确定纳入管理,数据库可实现本次骨折后2年档案构建、辅助宣教、智能随访、院内门诊电脑同屏显示等智能化管理功能。结论“骨质疏松性骨折数据库”拥有便捷的OF患者信息抓取功能,可实时了解相应管理的基础数据,可自动完成规定时间内设定管理的指导及提醒。该数据库有院内多源异构数据整合的专病数据库特点,为OF精准化、智能化、便捷化管理提供新的思路和有效工具。Objective To improve the quality of care and management efficiency for patients with osteoporotic fractures(OF),an automated“osteoporotic fracture database”within the hospital setting was developed which incorporated intelligent function modules for managing clinical workflows and was evaluated for its effectiveness in a real-world setting.Methods A closed-loop,hospital-based database designed for integrating heterogeneous data from multiple sources was constructed.The database seamlessly interfaces with existing hospital information systems including the hospital information system(HIS),picture archiving and communication system(PACS),and laboratory information system(LIS).Natural language processing(NLP)technology was implemented for automatic identification and integration of related information of the OF patients.A retrospective analysis was conducted to evaluate the intelligent management application of the database by data obtained from 12754 patients aged≥50 years with vertebral,hip,proximal humeral,or distal radial fractures who were admitted in Second Affiliated Hospital of Soochow University between June 2022 and June 2024.Results The database efficiently retrieved data from 12754 patients based on predefined criteria.It automatically collected and structured 407 data fields including patient demographics,fracture records from medical history and imaging examinations,laboratory test results,and real-time treatment plans.The database demonstrated intelligent capabilities in automatically identifying and categorizing osteoporosis-related data,such as fracture sites,bone mineral density(BMD)values,bone metabolism markers,and anti-osteoporosis medication use.Additionally,it enabled rapid and intelligent screening and combination of information related to in-hospital transfers,attending physician tracking,and history of multiple fractures.Once a patient was enrolled for management,the database automatically generated a two-year follow-up record,facilitated personalized education,provided intelligent fo

关 键 词：骨质疏松性骨折 再骨折 专病数据库 智能化管理 软件开发 

分 类 号：R319[医药卫生—基础医学]