Moher D, Fortin P, Jadad A R, Jüni P, Klassen T, Le Lorier J, Liberati A, Linde K, Penna A
Clinical Epidemiology Unit, Loeb Medical Research Institute, Ottawa, Ontario, Canada.
Lancet. 1996 Feb 10;347(8998):363-6. doi: 10.1016/s0140-6736(96)90538-3.
Lately, the number of systematic reviews published has increased substantially. Many systematic reviews exclude trials published in languages other than English. However, there is little empirical evidence to support this action. We looked for differences in the completeness of reporting between trials published in other languages and those published in English, to see whether the exclusion of trials published in other languages is justified.
We compared completeness of reporting, design characteristics, and analytical approaches of 133 randomised controlled trials (RCTs) published in English between 1989 and 1994 and 96 published in French, German, Italian, or Spanish during the same time. RCTs were identified by hand searching of journals (seven in English and six in the other languages).
We found no significant differences between trials published in English and other-language trials for any single item in the completeness of reporting scale (randomisation, double-blinding, withdrawals), or for the overall score (percentage of maximum possible score 51.0% for trials in English, 46.2% for trials in other languages; 95% CI for difference -1.1 to 10.5). Other-language trials were more likely than English-language trials to have adult participants, to use two or more interventions, and to compare two or more active treatments without an untreated control group. Trials in other languages were less likely to report a clearly prespecified primary outcome or any rationale for sample size estimation.
These results provide evidence for inclusion of all trial reports, irrespective of the language in which they are published, in systematic reviews. Their inclusion is likely to increase precision and may reduce systematic errors. We hope that our findings will prove useful to those developing guidelines and policies for the conduct of reporting of systematic reviews.
最近，发表的系统评价数量大幅增加。许多系统评价排除了非英文发表的试验。然而，几乎没有实证证据支持这一做法。我们研究了非英文发表的试验与英文发表的试验在报告完整性方面的差异，以确定排除非英文发表的试验是否合理。
我们比较了1989年至1994年间发表的133篇英文随机对照试验（RCT）和同期发表的96篇法语、德语、意大利语或西班牙语RCT在报告完整性、设计特征和分析方法方面的情况。通过手工检索期刊（7种英文期刊和6种其他语言期刊）来识别RCT。
我们发现，在报告完整性量表的任何单个项目（随机分组、双盲、撤稿）或总体得分方面，英文发表的试验与其他语言发表的试验之间没有显著差异（英文试验的最大可能得分百分比为51.0%，其他语言试验为46.2%；差异的95%置信区间为-1.1至10.5）。其他语言发表的试验比英文发表的试验更有可能纳入成年参与者，使用两种或更多干预措施，以及比较两种或更多活性治疗且无未治疗对照组。其他语言发表的试验不太可能报告明确预先指定的主要结局或样本量估计的任何理由。
这些结果为在系统评价中纳入所有试验报告提供了证据，无论其发表语言如何。纳入这些报告可能会提高精确性并可能减少系统误差。我们希望我们的研究结果将对制定系统评价报告指南和政策的人员有所帮助。
