Rodrigues S P, van Eck N J, Waltman L, Jansen F W
Department of Gynecology, Leiden University Medical Center, Leiden, The Netherlands.
BMJ Open. 2014 Mar 13;4(3):e004468. doi: 10.1136/bmjopen-2013-004468.
The amount of scientific literature available is often overwhelming, making it difficult for researchers to have a good overview of the literature and to see relations between different developments. Visualisation techniques based on bibliometric data are helpful in obtaining an overview of the literature on complex research topics, and have been applied here to the topic of patient safety (PS).
On the basis of title words and citation relations, publications in the period 2000-2010 related to PS were identified in the Scopus bibliographic database. A visualisation of the most frequently cited PS publications was produced based on direct and indirect citation relations between publications. Terms were extracted from titles and abstracts of the publications, and a visualisation of the most important terms was created. The main PS-related topics studied in the literature were identified using a technique for clustering publications and terms.
A total of 8480 publications were identified, of which the 1462 most frequently cited ones were included in the visualisation. The publications were clustered into 19 clusters, which were grouped into three categories: (1) magnitude of PS problems (42% of all included publications); (2) PS risk factors (31%) and (3) implementation of solutions (19%). In the visualisation of PS-related terms, five clusters were identified: (1) medication; (2) measuring harm; (3) PS culture; (4) physician; (5) training, education and communication. Both analysis at publication and term level indicate an increasing focus on risk factors.
A bibliometric visualisation approach makes it possible to analyse large amounts of literature. This approach is very useful for improving one's understanding of a complex research topic such as PS and for suggesting new research directions or alternative research priorities. For PS research, the approach suggests that more research on implementing PS improvement initiatives might be needed.
现有的科学文献数量往往多得令人难以招架,这使得研究人员很难全面了解文献情况并看清不同进展之间的关系。基于文献计量数据的可视化技术有助于获取复杂研究主题的文献综述,在此已应用于患者安全(PS)主题。
基于标题词和引用关系,在Scopus文献数据库中识别出2000 - 2010年期间与PS相关的出版物。基于出版物之间的直接和间接引用关系,对被引用最频繁的PS出版物进行可视化呈现。从出版物的标题和摘要中提取术语,并创建最重要术语的可视化。使用一种对出版物和术语进行聚类的技术,确定文献中研究的主要PS相关主题。
共识别出8480篇出版物,其中1462篇被引用最频繁的出版物被纳入可视化分析。这些出版物被聚类为19个簇,分为三类:(1)PS问题的严重程度(占所有纳入出版物的42%);(2)PS风险因素(31%)和(3)解决方案的实施(19%)。在PS相关术语的可视化中,识别出五个簇:(1)药物治疗;(2)衡量危害;(3)PS文化;(4)医生;(5)培训、教育与沟通。出版物层面和术语层面的分析均表明对风险因素的关注度在增加。
文献计量可视化方法能够分析大量文献。这种方法对于增进对诸如PS这样的复杂研究主题的理解以及提出新的研究方向或替代研究重点非常有用。对于PS研究,该方法表明可能需要更多关于实施PS改进举措的研究。