level 1
伴地莲阿斯顿
楼主
最近写了一篇meta分析，因为RCT实验数量太少，将RCT和队列研究同时纳入，再分亚组分别分析RCT和队列研究。进行证据等级评价的时候采用的是GRADE分级。现在的问题是，RCT和队列研究得出的结论虽然是一样的，但是不会进行evidence strength的评价。杂志编辑给的回复也不完全认同，有大神能帮我个忙么，不胜感激！！
以下是编辑的回复：
The evidence ratings completely overstate the quality of evidence. The authors need to refer to Guyatt et al 2011 https://www.jclinepi.com/article/S0895-4356(11)00206-X/fulltext
Where the confidence intervals remain wide, and especially when they remain wide and include the null result, only low quality evidence is possible as a rating. In my opinion the correct rating for every outcome is low quality evidence, because of either high inconsistency, high imprecision, high risk of bias, reliance on observational data or a combination of factors. It is also my opinion that long term rates of complications are highly important here, but are not assessed, and therefore directness is also reduced. The discussion needs to reflect these revised ratings.
有一篇是参考的别人发表的GRADE分级评价，可供参考，虽然不知道对不对
2020年02月03日 18点02分
1
以下是编辑的回复：
The evidence ratings completely overstate the quality of evidence. The authors need to refer to Guyatt et al 2011 https://www.jclinepi.com/article/S0895-4356(11)00206-X/fulltext
Where the confidence intervals remain wide, and especially when they remain wide and include the null result, only low quality evidence is possible as a rating. In my opinion the correct rating for every outcome is low quality evidence, because of either high inconsistency, high imprecision, high risk of bias, reliance on observational data or a combination of factors. It is also my opinion that long term rates of complications are highly important here, but are not assessed, and therefore directness is also reduced. The discussion needs to reflect these revised ratings.
有一篇是参考的别人发表的GRADE分级评价，可供参考，虽然不知道对不对