欢迎访问《中国循证儿科杂志》官方网站，今天是
English Version
中国循证儿科杂志

中国循证儿科杂志 ›› 2022, Vol. 17 ›› Issue (6): 469-474.DOI: 10.3969/j.issn.1673-5501.2022.06.012

• 论著 • 上一篇    下一篇

基于决策支持系统的新生儿转运给药知识库构建和应用效果验证

傅唯佳1,7,王颖雯1,7,张澜1,顾莺1,葛小玲1,王蓓蓓1,苏玲1,冯瑞2,曹云1,汪吉梅3,唐征4,刘江勤5,叶成杰1,王道洋6,汤梁峰1,蒋龙泉2,张帆1,郑如意1,周建国1,张晓波1
  

  1. 1 复旦大学附属儿科医院 上海，201102；2 复旦大学计算机科学技术学院 上海，200433；3 复旦大学附属妇产科医院 上海，200433；4 上海国际和平妇幼保健院 上海，200030；5 上海市第一妇婴保健院 上海，200040；6 上海市重大传染病和生物安全研究院，复旦大学公共卫生学院 上海，200032；7 共同第一作者

  • 收稿日期:2022-10-08 修回日期:2022-08-30 出版日期:2022-12-25 发布日期:2022-12-25
  • 通讯作者: 周建国；张晓波

Construction and validation of a decision support systembased knowledge base for neonatal transfer medication administration

FU Weijia1,7, WANG Yingwen1,7, ZHANG Lan1, GU Ying1,GE Xiaoling1, WANG Beibei1,SU Ling1, FENG Rui2, CAO Yun1, WANG Jimei3, TANG Zheng4, LIU Jiangqin5, YE Chengjie1, WANG Daoyang6, TANG Liangfeng1,1   JIANG Longquan2, ZHANG Fan1, ZHENG Ruyi1, ZHOU Jianguo1, ZHANG Xiaobo1    

  1. 1 Children's Hospital of Fudan University, Shanghai 201102,China; 2 School of Computer Science, Fudan University, Shanghai 200433,China; 3 The Obstetrics & Gynecology Hospital of Fudan University, Shanghai 200433,China; 4 The International Peace Maternity & Child Health Hospital, Shanghai 200030,China; 5 Shanghai First Maternity and Infant Hospital Tongji University School of Medicine, Shanghai 200040,China; 6 Shanghai Institute of Infectious Disease and Biosecurity, School of Public Health, Shanghai 200032, China; 7 Co-first author
  • Received:2022-10-08 Revised:2022-08-30 Online:2022-12-25 Published:2022-12-25
  • Contact: ZHOU Jianguo;ZHANG Xiaobo

PDF

176

补充材料

1

摘要： 背景 在5G+区块链建立的转出医院、医疗急救中心和接诊医院三位一体的危重新生儿急救转诊网络体系中，实现“上车即入院”，提升危重新生儿转运（NT）质量。 目的 构建多场景NT给药方案标准化知识库，开发多场景NT给药临床决策支持系统。 设计 质量改进研究。 方法 检索急危重NT途中给药管理的指南或共识，提取不同转运场景下药物应用证据，并通过专家共识构建NT给药知识库，基于决策支持系统帮助转运医生规范用药，分析其运行前（对照组）后（干预组）在上海市内NT药物使用情况。 主要结局指标 NT用药次数和种类。 结果 中英文数据库和其他资源系统共检索到149篇文献，去重、初筛和全文筛选后纳入了1篇专家共识，同时增加权威专著3部作为证据来源进入不同场景NT药物使用知识库的提取，共识专家组细化了低血糖场景设置（分为症状型和非症状型），对12种药物的推荐方案做了细化，NT不同场景药物使用知识库确定了16种药物。对照组30例，干预组61例，两组患儿性别、病种、胎龄差异均无统计学意义（P>0.05），出生体重差异有统计学意义（P=0.02）。对照组NT中使用了33次4种药物，干预组NT中使用了88次11种药物，其中由决策支持系统智能识别后触发弹窗提醒使用74次（85.9%），由医生主观判断选择场景后触发使用14次（14.1%）。干预组用药种类覆盖了对照组，对照组人均使用药物次数比为1.1(33/30)，较干预组［1.4（88/61）］差异有统计学意义( P=0.041)；对照组人均使用药物种类占比为25%（4/16），较干预组［62.5%（11/16）］差异有统计学意义（P=0.013）。 结论 构建的NT给药知识库及决策支持系统促进了NT中的规范用药和积极用药。

关键词: 新生儿, 院际转运, 程序化, 药物, 知识库, 决策支持系统

Abstract: Background A trinity of emergency referral networks consisting of transferring hospitals, medical emergency centers and receiving hospitals has been established based on 5G and blockchain so as to improve the quality of critical neonatal transfers by realizing admission on the ambulance. Objective To construct a standardized knowledge base for drug administration under multi-scenarios during neonatal transport (NT), followed by developing a NT drug administration clinical decision support system (CDSS). Design Quality improvement research. Methods Through expert consensus after retrieving guidelines or consensus on the drug administration during NT under emergent and critical conditions and extracting the evidence of drug administration under different transport scenarios, a knowledge base for NT drug administration was built to help doctors in transport team standardize drug administration by the CDSS and analyze the drug usage during NT in Shanghai before (control group) and after (intervention group) the application of the CDSS.Main outcome measuresNT drug types and the frequency of drug usage. Results A total of 149 articles were retrieved from Chinese, English literature databases and other resources. One expert consensus was included after duplication removal, primary screening, and full-text screening. Meanwhile, three authoritative monographs were added. The regimens of drug administration during NT were extracted from the above evidence sources. The consensus expert group refined the scenario settings of hypoglycemia for symptomatic and asymptomatic types and the regimens of 12 drugs administration. Finally, 16 drugs were confirmed in the knowledge base for NT drug administration. There were 30 cases in the control group and 61 cases in the intervention group. There was no significant difference in gender, disease type and gestational age between the two groups (P> 0.05), but there was significant difference in birth weight (P=0.02). Four drugs were used in the control group during NT for 33 times, and 11 drugs for 88 times in the intervention group, of which 74 (85.9%) were triggered by the CDSS and 14 (14.1%) by the subjective judgment of the doctors. The types of drugs used in the intervention group covered those in the control group. The mean of drug usage times in the control group was 1.1 (33/30), which was statistically significant compared with that of 1.4 in the intervention group (88/61). The proportion of drug usage per newborn in the control group was 25% (4/16), which was statistically significant compared with that of 62.5% in the intervention group (11/16) (P=0.031). Conclusion The construction of the knowledge base and CDSS for drug administration during NT has promoted standardized and on-demand drug administration.

Key words: Neonate, Inter-hospital transport, Procedural, Medication, Knowledge base, Decision support systems