摘要：

【目的】 调研评估我国主办的生物医学英文期刊的出版政策对研究开放和透明的支持度,探寻适合评估中国医学期刊透明度的工具。【方法】 采用两种工具（透明度和开放性评估工具、COS透明系数）对期刊支持开放性与透明度进行评价。通过查询Web of Science网站,获取2019年SCIE收录的我国主办的英文期刊及其影响因子,筛选获得影响因子大于3的生物医学类期刊,为Ⅰ组期刊;选取2019年“中国科技期刊卓越行动计划”申报期刊中影响因子小于3的生物医学英文类期刊,为Ⅱ组期刊。调研上述期刊的期刊政策是否声明国际医学期刊编辑委员会规定的期刊对披露利益冲突的要求以及对ORCID、伦理审批、作者贡献声明的要求。根据调研结果,结合信息熵加权法和专家打分法明确期刊出版政策体现的研究透明支持度的关键条目,构建适合中国英文期刊透明度现状的评价工具,并对其评价效果进行初步验证。【结果】 Ⅰ组期刊共33种,Ⅱ组期刊共31种,大多数期刊都要求对临床试验进行预注册（Ⅰ组：19/33;Ⅱ组：11/31）,并鼓励使用报告指南（Ⅰ组：16/33;Ⅱ组：12/31）。Ⅰ组、Ⅱ组分别有22种、16种期刊要求数据共享。Ⅰ组期刊COS透明系数中位数为2（IQR1）,Ⅱ组期刊为1（IQR1）。根据收集到的11位专家打分表及应用信息熵加权法分析研究结果,构建以报告指南、伦理审批、预注册、利益冲突声明、作者贡献声明和数据共享声明为构成要素的评价工具,利用该工具重新评估所调研的64种期刊的出版政策对研究透明度的支持度,其与期刊影响因子之间呈弱相关（r=0.435,P=0.003）。【结论】 我国期刊出版政策对研究透明度的支持度亟需提升。期刊透明度相关评价体系可作为医学期刊质量,或者医学期刊论文质量评价体系的补充。

关键词: 透明度, 开放科学数据, 期刊评价体系

Abstract:

[Purposes] This paper aims to investigate the requirement of research openness and transparency in the publishing policies of English medical journals in China and to explore appropriate tools for evaluating the transparency of Chinese medical journals.[Methods] Two methods （transparency and openness evaluation tool and COS transparency coefficient） were used to evaluate the support of research openness and transparency by journals. English journals of China indexed by SCIE in 2019 and their impact factors were retrieved from Web of Science, and the biomedical journals with impact factor >3 were selected and categorized into group Ⅰ. Biomedical English journals which were included in the Excellence Action Plan for China STM Journals in 2019 were taken as group Ⅱ. In addition, we investigated whether the polices of the above journals included the requirements of the ICMJE for disclosure of conflicts of interest and for open researcher and contributor identification numbers （ORCID）, ethical approval, and author contribution statement. According to the survey results, the key items on research transparency in journal publishing policies were identified by entropy weight method and expert scoring. On this basis, an evaluation tool for the transparency of English journals in China was developed and its effect was tested. [Findings] A total of 33 journals were included in group Ⅰ and 31 in group Ⅱ. Most of the journals required preregistration of clinical trials （19/33 in group Ⅰ, 11/31 in group Ⅱ） and encouraged the use of reporting guidelines （16/33 in group Ⅰ, 12/31 in group Ⅱ）. A total of 22 journals in group Ⅰ and 16 in group Ⅱ required data sharing. The median transparency coefficient of COS in group Ⅰ was 2 （IQR1）, and that in group Ⅱ was 1 （IQR1）. According to the scores from 11 experts, the entropy weight method was used to analyze the research results. An evaluation tool consisting of report guidelines, ethical approval, preregistration, statement of conflict of interest, statement of author contribution, and statement of data sharing, was constructed. The tool was used to reassess the degree to which publishing policies in 64 journals surveyed in this study supported research transparency, and the result showed that the support was in weak correlation with journal impact factor （r=0.435, P=0.003）.[Conclusions] The publishing policies of journals should further emphasize research transparency. The evaluation system for journal transparency can be used as a supplement to the quality evaluation system of medical journals, or in other words, the quality evaluation system of medical articles.

Key words: Transparency, Open scientific data, Journal evaluation system