MIMIC数据库研究现状与热点分析

Research Status and Hot Spot Analysis of MIMIC Database

目的检索MIMIC数据库研究的相关文献,分析当前研究现状与热点。方法在PubMed数据库中下载2010年1月-2020年12月有关MIMIC数据库的相关文献题录,分析文献的发表年份、来源期刊及期刊所属国家、第一作者;使用BICOMB 2.0统计主要主题词出现的频次,截取频次≥4次的高频主要主题词形成词篇矩阵,运用SPSS 22.0进行系统聚类分析。结果关于MIMIC数据库的发文量共334篇,共有91种期刊发表了相关文献,期刊来源国家排名前3的是美国、英国、荷兰;第一作者主要分布在美国和加拿大;高频主题词有30个,通过共词聚类分析获得6个研究热点:MIMIC数据库介绍或数据处理方法的研究,ICU患者预后、死亡率预测分析,对ICU患者基本生命体征、护理信息的研究,探究某些因素是否是某些疾病的影响因子,范围较广的其他类研究,预测某种疾病的发病率或死亡率。结论基于MIMIC数据库已经开展了很多项临床数据挖掘研究,针对护理病程记录、护理相关信息的研究还较少,今后可根据护理领域数据开展更多研究。借鉴MIMIC数据库运行机制构建符合我国患者的临床医学数据集,可有效辅助临床决策。

Objective To search the related literature of MIMIC database and analyze the current research status and hotspots.Methods The relevant literature titles of MIMIC database from January 2010 to December 2020 were downloaded from PubMed database,and the publication year,source journals,country and first author of the literature were analyzed.BICOMB 2.0 was used to calculate the frequency of the occurrence of the main topic words,and the high-frequency main topic words with the frequency≥4 times were intercepted to form the discourse matrix.SPSS 22.0 was used for systematic clustering analysis.Results A total of 334 articles were published in the MIMIC database,and related articles were published in 91 journals.The top 3 journal source countries were the United States,the United Kingdom,and the Netherlands;the first authors were mainly located in the United States and Canada.There were 30 high-frequency keywords,and six research hotspots were obtained through co-word clustering analysis:the introduction of MIMIC database or the research on data processing methods,the prediction and analysis of prognosis and mortality of ICU patients,the research on basic vital signs and nursing information of ICU patients,and the exploration of whether certain factors are the influencing factors of certain diseases,other types of research with a wide range can predict the incidence or mortality of certain diseases.Conclusion Many clinical data mining studies have been carried out based on the MIMIC database,and there are few studies on nursing course records,nursing-related information,etc.,and more studies can be carried out based on data in the nursing field in the future.Use the MIMIC database operating mechanism for reference to construct clinical medical data sets that are in line with patients in my country,and effectively assist clinical decision-making.