论文临床影响力评价及预测指标的实证研究--基于诺贝尔生理学或医学奖获得者成果的分析  

Exploring the Clinical Impact Predictors of Papers-Based on the Empirical Analysis of Publications from Nobel PrizeWinners in Physiology or Medicine

作　　者：陈斯斯[1] 刘春丽[1] Chen Sisi;Liu Chunli(Library of China Medical University,Shenyang 110122)

机构地区：[1]中国医科大学图书馆,沈阳110122

出　　处：《情报学报》2022年第2期142-154,共13页Journal of the China Society for Scientific and Technical Information

基　　金：2020年度辽宁省图书馆学会研究课题“基于开放引文集的医学生物学论文临床转化能力评价指标研究”(2020tsgxhybkt-024);辽宁省教育厅2020年度科学研究经费项目“COVID-19科研成果的学术影响力、社会关注度及临床转化潜力评价与预测研究”(QNRW2020004)。

摘　　要：在重大突发公共卫生事件的背景下,科技论文的临床应用价值被提到前所未有的重要位置。但如何评价这种类型的影响力以及有哪些有效指标仍需要深入挖掘与探索。引文桂冠奖是基于被引频次的诺贝尔生理学或医学奖的预测方法,而未被其预测却最终获得诺贝尔奖的原因可能是传统被引指标无法探测到论文的潜在临床影响力。本文引入美国NIH(National Institutes of Health)提出的论文临床转化潜力近似值指标(approximate potential to translate scores,APT),选择诺贝尔生理学或医学奖得主的论文集为样本,比较被引文桂冠奖预测和未被预测两组作者论文集的总被引次数、加权RCR(relative citation ratio)、被临床论文引用次数、APT均值、Human均值、Animal均值、Mol/Cell(Molecular/Cellular)均值这7项指标,以及转化力三角形模型的差异与指标间的相关性。被预测获奖和未被预测获奖两组论文的总被引次数、加权RCR、Mol/Cell均值有显著差异,未被预测组的Human均值与Animal均值及中位数均高于被预测组。从三角形模型上来看,被预测组的Mol/Cell均值较为明显,未被预测组的Human均值和Ani‐mal均值较明显。从相关分析结果可以看出,Mol/Cell均值与总被引次数显著正相关,而Animal均值与总被引次数显著负相关(P<0.05),Human均值与总被引次数相关性不显著。引文桂冠奖中使用的被引次数无法评价或预测临床转化方面的影响力,对分子、细胞生物学层面的基础研究影响力测度更为敏感,而对动物实验和临床研究影响力缺乏敏感性。而APT指标可较好地测度论文未来临床转化潜力,可考虑纳入科技论文评价指标体系,以弥补单纯基于引文分析指标的不足。Under the context of major public health emergencies,the clinical application value of scientific papers is being given importance.However,the evaluation this type of influence and the indicators of effectiveness still need to be ex-plored.The Citation Laureate is a method used in predicting the winner of the Nobel Prize in Physiology and Medicine based on the frequency of citations.The reason why the Nobel Prize was finally won without its prediction may be because the traditional citation indicators cannot detect the potential clinical impact of the study.This study introduces the Approxi-mate Potential to Translate scores(APT)proposed by NIH in the United States,uses Nobel Prize winners in Physiology or Medicine as the sample,and compares between research collections by predicted authors and unpredicted authors.Seven indicators:Total Citations,Weighted RCR,Cited By Clin,Avg.APT,Avg.Human,Avg.Animal,and Avg.Mol/Cell are de-termined and their differences are evaluated using the translation triangle model,and correlations between those indicators are identified.Total Citations,Weighted RCR,and Avg.Mol/Cell significantly differ between predicted and unpredicted pa-pers.The mean and median values of Avg.Human and Avg.Animal of the unpredicted group are higher than those of the predicted group.From the perspective of the triangle model,the molecular/cell characteristics of the predicted group are highlighted,whereas Avg.Human and Avg.Animal indicators of the unpredicted group are more obvious.The results of correlation analysis show that Avg.Mol/Cell is significantly positively correlated with Total Citations,while Avg.Animal is significantly negatively correlated with Total Citations,and Avg.Human is not significantly correlated with Total Cita-tions.The number of citations used in the Citation Laureate Award cannot evaluate or predict the impact of clinical transla-tion,and is more sensitive to the measurement of the impact of basic research on molecular and cell biology;however,it lacks sensitivity to the impact

关 键 词：临床转化 临床影响力 iCite 预测 

分 类 号：G353.1[文化科学—情报学] R-05[医药卫生]