Mills Edward J, Wu Ping, Gagnier Joel, Devereaux P J
Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University, HSC-2C12, 1200 Main Street West, Hamilton, ON, Canada, L8N 3Z5.
Contemp Clin Trials. 2005 Aug;26(4):480-7. doi: 10.1016/j.cct.2005.02.008. Epub 2005 Mar 31.
If randomized controlled trials (RCTs) are to fulfill their potential to allow health care providers to make informed inferences about the validity of the trials upon which they base their clinical practice, then the reporting of key methodological factors must improve. The revised Consolidated Standards for Reporting Trials (CONSORT) were developed to improve the suboptimal reporting of RCTs. We aimed to determine the extent to which RCTs report key methodological factors included in the revised CONSORT recommendations following publication of these recommendations. We conducted an observational study of RCTs published between July 2002 and June 2003 in 5 leading general medicine journals. We determined the proportion and 95% confidence intervals of RCTs that reported sequence generation, allocation concealment, randomization implementation, blinding status of participants, health care providers, outcome assessors, and data analysts, sample size justification, method of analysis (e.g. intention-to-treat), and a participant flow diagram. We included a total of 253 RCTs. RCTs reported: sequence generation 80% (95% CI 75-80%), allocation concealment 48% (95% CI 42-54%), randomization implementation 55% (95% CI 49-61%), blinding status of participants 40% (34-46%), health care providers 17% (95% CI 13-22%), outcome assessors 47% (95% CI 41-53%), data analysts 15% (10-19%), sample size justification 83% (95% CI 78-87%), method of analysis 87% (95% CI 83-91%)), and participant flow diagrams 86% (95% CI 82-90%). Although reporting of some CONSORT recommendations is high, reporting of several essential recommendations remains suboptimal. Authors need to comply with and journals need to enforce reporting of the revised CONSORT recommendations.
如果随机对照试验（RCT）要充分发挥其潜力，使医疗保健提供者能够对作为临床实践依据的试验有效性做出明智推断，那么关键方法学因素的报告必须加以改进。修订后的《试验报告统一标准》（CONSORT）旨在改善随机对照试验报告欠佳的情况。我们的目的是确定在这些建议发布后，随机对照试验报告修订后的CONSORT建议中所包含关键方法学因素的程度。我们对2002年7月至2003年6月在5种主要的普通医学期刊上发表的随机对照试验进行了一项观察性研究。我们确定了报告序列产生、分配隐藏、随机化实施、参与者、医疗保健提供者、结果评估者和数据分析人员的盲法状态、样本量合理性、分析方法（如意向性分析）以及参与者流程图的随机对照试验的比例和95%置信区间。我们总共纳入了253项随机对照试验。随机对照试验报告的情况如下：序列产生80%（95%置信区间75 - 80%），分配隐藏48%（95%置信区间42 - 54%），随机化实施55%（95%置信区间49 - 61%），参与者盲法状态40%（34 - 46%），医疗保健提供者17%（95%置信区间13 - 22%），结果评估者47%（95%置信区间41 - 53%），数据分析人员15%（10 - 19%），样本量合理性83%（95%置信区间78 - 87%），分析方法87%（95%置信区间83 - 91%），以及参与者流程图86%（95%置信区间82 - 90%）。尽管对一些CONSORT建议的报告率较高，但对一些重要建议的报告仍然欠佳。作者需要遵守，期刊需要强制要求报告修订后的CONSORT建议。
