[AF-FORUM] 117E 关于随机试验. Dr. Perez Riera

[AF Symposium]

亲爱的Mauricio Rondon医生
我同意主观倾向确实存在，但随机试验的重要性也不 
能过分强调。能够发表与否不仅取决于发现本身，还 
取决于编辑和审稿的过程。在编辑或者审稿过程中很 
多因素，包括审稿人的主观倾向都需要考虑。仅有63% 
的随机或者对照实验的摘要能够最后全文发表。阳性 
结果比非阳性结果的发表纪律要高（1）。随机试验 
往往对于得出一个结论是必要的，明确的，而且可防 
御的。缺失的一些因素可以通过实验前一些依赖因素 
的整合而替代。实验前以及试验的数据可以通过负性 
三项模型所分析（2）。在原始健康数据的随机试验 
分析中方法学的质量是不同的，而且报告的方式也需 
要提高。Eldrige等（3）整理集群试验的系统注册，进 
行考克兰对照注册研究。这些作者还收集一些会议上 
的和英国国家研究注册中心未发表的实验数据。他们 
依据一些清单，以及研究者研究时遇见的一些问题评 
估方法学的质量，并观察集群多个实验（而非单个的 
随机试验）在分析时到何种程度会减少权重。作者研 
究了367个实验报告。许多实验报告不止一次。他们分 
析了自1997到2000年健康领域152个独立的随机试验，并 
简要描述了47个2000年底还没有发表的实验。发现设计 
和分析的质量是有差别的。在发表的实验中，20%的计 
算方法在聚类研究中是可靠的，59%的分析方法在聚类 
研究中是可靠的。未发表的或最近的实验相对质量更 
高一些。报告的质量在那些经常报告这种试验的杂志 
中质量会更好一些，许多研究者报告对于坚持设计方 
案，招募病例以及干预方式方面会遇到一些困难。科 
学训练本身应该使得这些研究人员坚守伦理，这不仅 
是因为这样可以使得他们更好的为全人类负责，也使 
得他们自己成为一个真正更好的科学研究者（4）

参考文献
1)        Scherer R, Langenberg P, von Elm E, et al. Full publication
of results initially presented in abstracts. Cochrane Database Syst
Rev. 2007;(2):MR000005.
2)         Duffy SW, South MC, Day NE. Cluster randomization in large
public health trials: the importance of antecedent data. Stat Med.
1992; 11:307-316.
3)        Eldrige SM, Ashby D, Feder GS, et al. Lessons for cluster
randomized trials in the twenty-first century: a systematic review of
trials in primary care. Clin Trials. 2004;1:80-90.
4)        De Melo-MatinI, Intermann KK. Can ethical reasoning
contribute to better epidemiology? A case study in research on racial
health disparities. Eur J Epidemiol. 2007 Feb 15; [Epub ahead of print]

All the best for all
Andrés Ricardo Pérez Riera MD
Chief of Electro-Vectocardiology Sector of the Discipline of
Cardiology,
ABC Faculty of Medicine (FMABC), Foundation of ABC (FUABC) - Santo
André 圣保罗 – 巴西.
riera在uol.com.br

-----------------------------------------------------
Re: [AF-FORUM] 117E About randomization. Dr. Perez Riera

Dear Dr Mauricio Rondon:
I agree with the real presence of bias, but the importance of  
randomization cannot be over stressed. The ability to publish will be  
dependent not only on the significance of the findings, but also on  
the editorial staff of the journal and the peer review process. The  
manner in which the peer review process can be influenced by a  
journal's staff, as well as by the potential bias of reviewers, needs  
appreciation. Only 63% of results from abstracts describing  
randomized or controlled clinical trials are published in full.  
'Positive' results are more frequently published than not 'positive'  
results (1). Randomization is necessary for conclusions drawn from  
the experiment to be correct, unambiguous and defensible. The lost  
power can largely be replaced by incorporating information on the  
dependent variable, within clusters, before the start of the trial.  
The pretrial and trial data can be analyzed by negative trinomial  
models (2). Methodologic quality of cluster randomized trials in  
primary health care is variable and reporting needs improvement.  
Eldrige et al (3) conducted a systematic review of cluster randomized  
trials in primary health care, searching the Cochrane Controlled  
Trials Register. The authors also searched for unpublished trials in  
conference proceedings, and the UK National Research Register. They  
assess methodological quality using a checklist, articulate problems  
facing investigators conducting these trials, and examine the extent  
to which carrying out a cluster randomized trial (as opposed to an  
individually randomized trial) in primary care may reduce power. The  
authors found 367 trial reports. Many trials were reported more than  
once. They characterize 152 independent cluster randomized trials in  
primary health care published between 1997 and 2000, and briefly  
describe 47 trials unpublished at December 2000. The quality of  
design and analysis was variable. Of published trials reporting  
sample size calculations 20% accounted for clustering in these  
calculations, 59% of published trials accounted for clustering in  
analyses. Unpublished trials were more recent and of higher quality.
Reporting quality was better in journals reporting more cluster  
randomized trials. Many trial investigators reported problems with  
adherence to protocol, recruitment and type of intervention.  
Scientific training should prepare scientists to engage in ethical  
reasoning not only because it will make them more responsible human
beings, but also because it will make them better scientists (4).

References

1)        Scherer R, Langenberg P, von Elm E, et al. Full publication
of results initially presented in abstracts. Cochrane Database Syst
Rev. 2007;(2):MR000005.
2)         Duffy SW, South MC, Day NE. Cluster randomization in large
public health trials: the importance of antecedent data. Stat Med.
1992; 11:307-316.
3)        Eldrige SM, Ashby D, Feder GS, et al. Lessons for cluster
randomized trials in the twenty-first century: a systematic review of
trials in primary care. Clin Trials. 2004;1:80-90.
4)        De Melo-MatinI, Intermann KK. Can ethical reasoning
contribute to better epidemiology? A case study in research on racial
health disparities. Eur J Epidemiol. 2007 Feb 15; [Epub ahead of print]

All the best for all
Andrés Ricardo Pérez Riera MD
Chief of Electro-Vectocardiology Sector of the Discipline of
Cardiology,
ABC Faculty of Medicine (FMABC), Foundation of ABC (FUABC) - Santo
André São
Paulo - Brazil.
riera在uol.com.br


--
Dr. Sergio Dubner
President of Scientific Committee

Dr. Edgardo Schapachnik
President of Steering Committee