School of Nursing, University of Auckland, Auckland, New Zealand; National Institute for Health Innovation, University of Auckland, Auckland, New Zealand.
School of Nursing, University of Auckland, Auckland, New Zealand; School of Population Health, University of Auckland, Auckland, New Zealand.
Int J Nurs Stud. 2015 Jun;52(6):1071-9. doi: 10.1016/j.ijnurstu.2014.11.008. Epub 2014 Nov 18.
To establish the reporting quality of trials published in leading nursing journals and investigate associations between CONSORT Statement or trial registration endorsment and reporting of design elements.
The top 15 nursing journals were searched using Medline for randomised controlled trials published in 2012. Journals were categorised as CONSORT and trial registration promoting based on requirements of submitting authors or the journal's webpage as at January 2014. Data on sequence generation, allocation concealment, follow up, blinding, baseline equivalence and sample size calculation were extracted by one author and independently verified by the second author against source data.
Seven journals were CONSORT promoting and three of these journals were also trial registration promoting. 114 citations were identified and 83 were randomised controlled trials. Eighteen trials (21.7%) were registered and those published in trial registration promoting journals were more likely to be registered (RR 2.64 95%CI 1.14-6.09). We assessed 68.7% of trials to be low risk of bias for sequence generation, 20.5% for allocation concealment, 38.6% for blinding, 55.4% for completeness of follow up and 79.5% for baseline equivalence. Trials published in CONSORT promoting journals were more likely to be at low risk of bias for blinding (RR 2.33, 95%CI 1.01-5.34) and completeness of follow up (RR 1.77, 95%CI 1.02-3.10), but journal endorsement of the CONSORT Statement or trial registration otherwise had no significant effect. Trials published in CONSORT and trial registration promoting journals were more likely to have high quality sample size calculations (RR 2.91, 95%CI 1.18-7.19 and RR 1.69, 95%CI 1.08-2.64, respectively).
Simple endorsement of the CONSORT Statement and trials registration is insufficient action to encourage improvement of the quality of trial reporting across the most important of trial design elements.
评估发表在顶级护理期刊上的试验报告质量，并研究 CONSORT 声明或试验注册认可与设计要素报告之间的关联。
使用 Medline 对 2012 年发表的随机对照试验进行了 15 种顶级护理期刊的搜索。根据提交作者的要求或 2014 年 1 月的期刊网页，将期刊分类为 CONSORT 和试验注册推广。一位作者提取了关于随机序列生成、分配隐藏、随访、盲法、基线等效性和样本量计算的数据，并由第二位作者独立地根据原始数据进行验证。
有 7 种期刊是 CONSORT 推广期刊，其中 3 种也是试验注册推广期刊。共确定了 114 个引文，其中 83 个是随机对照试验。有 18 个试验（21.7%）进行了注册，在试验注册推广期刊上发表的试验更有可能进行注册（RR 2.64，95%CI 1.14-6.09）。我们评估了 68.7%的试验在随机序列生成方面为低偏倚风险，20.5%的试验在分配隐藏方面为低偏倚风险，38.6%的试验在盲法方面为低偏倚风险，55.4%的试验在随访完整性方面为低偏倚风险，79.5%的试验在基线等效性方面为低偏倚风险。在 CONSORT 推广期刊上发表的试验更有可能在盲法（RR 2.33，95%CI 1.01-5.34）和随访完整性（RR 1.77，95%CI 1.02-3.10）方面处于低偏倚风险，而 CONSORT 声明或试验注册的期刊认可对其他方面没有显著影响。在 CONSORT 和试验注册推广期刊上发表的试验更有可能具有高质量的样本量计算（RR 2.91，95%CI 1.18-7.19 和 RR 1.69，95%CI 1.08-2.64）。
简单地认可 CONSORT 声明和试验注册不足以鼓励改进最重要的试验设计要素的试验报告质量。
