情报科学 ›› 2024, Vol. 42 ›› Issue (8): 22-33.

• 理论研究 • 上一篇    下一篇

TQM视角下医疗健康大数据质量影响因素研究 ——基于模糊DEMATEL-ISM方法的实证分析

  

  • 出版日期:2024-08-01 发布日期:2024-11-05

  • Online:2024-08-01 Published:2024-11-05

摘要: 【目的/意义】本研究充分考虑医疗健康大数据特点,以医疗健康大数据质量为研究对象识别其影响因素, 厘清因素之间层级和结构关系,推动国家医疗健康大数据战略发展。【方法/过程】基于全面质量管理(TQM)理论, 通过文献分析法识别出医疗健康大数据质量影响因素,并运用模糊决策试验与评价试验(Fuzzy-DEMATEL)和解 释结构模型(ISM)分析其层级及内在联系。【结果/结论】研究发现,医疗健康大数据质量受5个维度17个因素的影 响,包括相关人员职业素养、利益相关方协同、信息系统 3个最直接的影响因素,供需环境、安全环境等 9个连接底 层和顶层的枢纽因素,政治环境、文化环境等5个根源因素。【创新/局限】本研究基于全面质量管理理论,从环境、组 织、标准、信息技术和数据属性五个层面系统地分析医疗健康大数据质量影响因素,具有一定创新性。不足之处在 于本研究未将已识别的因素权重量化,后续可通过相关权重的计算方法,具体测算各个影响因素的作用大小。

Abstract: 【Purpose/significance】This study fully considers the characteristics of medical and health big data, identifies the influenc⁃ ing factors of medical and health big data quality as the research object, clarifies the hierarchical and structural relationships between factors, and promotes the development of the national medical and health big data strategy.【Method/process】Based on the theory of Total Quality Management (TQM), the factors affecting medical and health big data quality were identified by literature analysis, and the hierarchical relationship and internal relationship were analyzed by Fuzzy-DEMATEL and ISM.【Result/conclusion】The study found that medical and health big data quality is affected by 17 factors in 5 dimensions, among which there are 3 most direct influenc⁃ ing factors, including the professional quality of relevant personnel, stakeholder collaboration and information system, 9 pivot factors such as the supply and demand environment and the security environment connecting the bottom and top relationships, and 5 root fac⁃ tors such as political environment and cultural environment【. Innovation/limitation】Based on the TQM theory, this study systematically analyzes the factors affecting medical and health big data quality from the five levels of environment, organization, standards, informa⁃ tion technology and data attributes, which is innovative to some extent. The limitation is that the weight of the identified factors is not quantified in this study, and the role of each influencing factor can be specifically measured by the relevant weight calculation method in the future.