2011-2013年国内期刊发表的帕金森病随机对照研究文献质量评价

Assessment of the quality of randomized controlled trials of Parkinson disease published in China from 2011 to 2013

目的了解国内期刊发表的有关帕金森病（PD）治疗的随机对照研究（RCT）的文献质量。方法采用CONSORT 2010声明标准和Jadad量表评分对2011-01—2013—12期间国内期刊发表的PD相关RCT研究报告的真实性和有效性进行评价。结果共纳入RCT报告256篇，其中2011年61篇，2012年98篇，2013年97篇。这3年中文献在描述试验设计（依次为11．5％、8．2％、21．6％）、解释中期分析情况和终止试验规则（依次为8．2％、2．0％、1．0％）、描述各组退组和剔除的人数及原因（依次为14．8％、5．1％、15．5％）、总结各组的主要和次要结局结果（依次为83．6％、94．9％、100．0％）报告比例上差异有统计学意义（均P〈0．05）。256篇文献中，253篇（98．8％）有结构式摘要，252篇（98．4％）阐述背景和原理，231篇（90．2％）有参加者合格标准，251篇（98．0％）描述干预措施，241篇（94．1％）总结主要、次要结局指标；仅有3篇（1．2％）文题提示RCT，36篇（14．1％）描述试验设计，2篇（0．8％）明确样本量的确定，63篇（24．6％）描述随机化方法，6篇（2．3％）描述分配隐蔽，18篇（7．0％）描述盲法，35篇（13．7％）用表格描述基线资料，45篇（17．6％）描述试验的局限性。3年中Jadad评分为3～5分的高质量文献分别为6篇（9．8％）、5篇（5．1％）和11篇（11．3％）。结论本组RCT文献的方法学描述不足，高质量文献较少，还需进一步推广和采用CONSORT声明和Jadad量表，使这类文献真实、可靠、有效。

Objective To assess the quality of randomized controlled trials （RCTs） for the treatment of Parkinson disease published in China from 2011 to 2013. Methods The authenticity and validity of those literatures were evaluated according to the Consolidated Standards of Reporting Trials （CONSORT） 2010 statement and Jadad scale. Results A total of 256 RCTs were included, including 61 in 2011, 98 in 2012, and 97 in 2013. There were significant differences in description of trial design （11.5% vs. 8.2% vs. 21.6%）, explanation of any interim analyses and stopping guidelines （8.2% vs. 2.0% vs. 1.0%）, losses and exclusions in each group together with reasons （14.8% vs. 5.1% vs. 15.5%）, and results for each group （83.6% vs. 94.9% vs. 100.0%, all P〈0.05）. In all the 256 literatures, 253 （98.8%） had structured abstracts, 252 （98.4%） described scientific background and explanation of rationale, 231 （90.2 %） had eligibility criteria for participants, 251 （98.0%） described interventions for each group, 241 （94.1%） summarized primary and secondary outcome measures. However, only 3 （1.2%） studies identified themselves as randomized trials in the title, 36 （14.1%） described trial design, 2 （0.8%） described how sample size was determined, 63 （24.6 %） described method used to generate the random allocation sequence, 6 （2.3 %） described allocation concealment mechanism, 18 （7.0 %） described blinding, 35 （13.7%） used a table showing baseline characteristics, 45 （17.6%） described trial limitations. There were only 6 （9.8%）, 5 （5.1%）, and 11 （ 11.3 %） high-quality literatures based on Jadad scales （≥3 points） in 2011, 2012 and 2013, respectively. Conclusions The methodology description is not sufficient and high-quality literatures are scarce in these RCTs, CONSORT statement and Jadad scale should be promoted and used for improving the authenticity and validity of RCT literatures.