基于多因素时间序列的医院科研绩效预测模型构建与应用  

Construction and Application of Scientific Research Performance Prediction Model in Hospital Based on Multifactor Time Series

作　　者：周争[1] 石陆华 汪洁滢 陆秉炜 杨晓秋 仓艺倩[1] 丁文彬[1] 胡丹 王青[1] 王争[1] 董菡珺[1] 顾乐怡[1] 戴慧莉[1] Zhou Zheng;Shi Luhua;Wang Jieying(Renji Hospital Affiliated to Shanghai Jiaotong University School of Medicine,Shanghai,200127,China;不详)

机构地区：[1]上海交通大学医学院附属仁济医院,上海200127 [2]上海交通大学数学科学学院,上海200240

出　　处：《中国医院管理》2022年第10期55-59,共5页Chinese Hospital Management

基　　金：上海申康医院发展中心临床科技创新项目(SHDC12017608)。

摘　　要：目的 构建一种科学、准确的科研绩效预测模型,为医院科研绩效评价提供决策依据。方法 基于上海交通大学医学院附属仁济医院2007—2018年的科研投入和产出数据,分析科研产出指标的影响因素。在此基础上,考虑科研投入转化为产出的时滞效应,采用滑窗方法对原始数据进行加权处理,构建多因素时间序列预测模型。结果 硕士研究生和博士研究生数量是科研产出指标的关键因素。科研项目经费对SCI论文数和影响因子有贡献,反之,SCI论文数和影响因子也能帮助提升科研项目数和经费。SCI论文影响因子对科研奖励数有较大的贡献;预测模型能较好地反映SCI论文数和影响因子、科研项目数和经费,但对于科研奖励数的预测能力不足。结论 在基于历史数据的时间序列模型中引入加权移动平均法构建科研绩效预测模型,使得科研绩效评价趋于合理、更具可靠性和预测性。Objective Constructing a scientific and accurate prediction model of scientific research performance to provide a decision-making basis for hospital scientific research performance evaluation. Methods Based on the scientific research input and output data of Renji Hospital Affiliated to Shanghai Jiaotong University School of Medicine from2007 to 2018, the influencing factors of the scientific research output index were analyzed. On this basis,considering the time lag effect of transforming scientific research input into output,the sliding window method is used to weight the original data to build multifactor time series,prediction model. Results The number of graduate students is the key factor of scientific research output index. The number and influencing factors of SCI papers are influenced by the funding of scientific research projects. On the contrary,the number and influencing factors of SCI papers can also help to increase the number and funding of research projects. The influence factors of SCI papers have a great contribution to the number of scientific research awards. The prediction model can better reflect the number and impact factors of SCI paper,the number and funding of scientific research projects,but the prediction ability for the number of scientific research awards is insufficient. Conclusion In the time series model based on historical data, the weighted moving average method is introduced to build the scientific research performance prediction model,which makes the scientific research performance evaluation more reasonable,more reliable and predictable.

关 键 词：医院 科研绩效 预测模型 

分 类 号：R197.323.6[医药卫生—卫生事业管理]