临床试验报告的统一标准声明评价2004-2008年国内五种精神科期刊随机对照试验文献报告质量

Quality assessment of the reports of randomized controlled trials published in 5 journals of psychiatry in China from 2004 to 2008 by Consolidated Standards of Reporting Trials Statement

目的运用临床试验报告的统一标准（CONSORT）声明评价2004-2008年国内5种精神科专业期刊发表的随机对照试验（RCT）文献的方法学质量。方法以“随机、对照”为关键词，计算机检索维普中文电子数据库中中华精神科杂志等5种精神科期刊刊载的可以获得全文的RCT文献，依据CONSORT声明修订版和其他补充评价方法，对文献质量进行综合评价。结果共纳入518篇RCT文献，占总发表文献数的8．24％。纳入文献在受试者纳入排除标准（92．1％）、干预（99．4％）、统计学描述（90．2％）、受试者募集（94．2％）、基线资料（98．1％）、结果（97．7％）和不良事件（87．8％）等方面报告较为详细，但在文题摘要（67．8％）和解释（77．4％）方面报告欠详尽，而有关RCT重要方法学，如样本量估算（0．4％）、随机序列产生（9．8％）、分配隐藏（1．4％）、随机分配的实施（1．5％）、主次要结局指标的确定（3．9％）、盲法的实施（14．7％）、意向性治疗分析（6．0％）和流程图的使用（0．0％）等方面报告不详尽或缺失。结论国内精神科期刊发表的临床随机对照试验文献报告质量与CONSORT声明标准差距较大，建议今后在精神科期刊中推广并采用CONSORT声明标准。

Objective To explore the quality of methodology about the reports of randomized controlled trials （RCTs） in 5 journals of psychiatry in China from 2004 to 2008 by CONSORT statement. Methods The reporting of randomized controlled trials related to psychiatry in five Chinese psychiatric academic journals were searched through Chongqing VIP full-text database （VIP）. According to the revised CONSORT statement and other addition methods of appraisal, the quality of randomized controlled trials was assessed. Results 518 studies, which account for 8.24% of all the articles were identified. Regarding to the criteria of inclusion and exclusion criteria about subjects （ 92. 1% ）, intervention （99. 4% ）, statistical description （ 90. 2% ）, recruitment （ 94. 2% ）, baseline data （ 98. 1% ）, results （97. 7% ） and adverse events（87.8% ）, most of the studies reported in details. But the titles and abstracts （67.8%）, and discussions （77.4%） , were reported less focused contents in detail. However, as to the most important methodology of randomized controlled trails was poorly reported or was not reported, for instance, the estimation of sample size （0.4%）, the sequence generation of randomization （9. 8% ）, allocation concealment （ 1.4% ） , implementation of randomization （ 1.5% ） , primary and secondary outcomes （3.9%） , blinding（ 14. 7% ）, analysis of the intention to treatment （6. 0% ）, and application of the flowchart （0. 0% ） , much content needed to improve. Conclusion The quality of the reports of RCTs in academic journals of psychiatry in China was still poor, and the CONSORT statement should be widely used in reports of RCTs.