I'm performing a systematic review and I don't understand how to assess/report the risk of bias in my search? I entered the GRADE website and couldn't find any formal recommendation.
Would appreciate your help!
Best,
The Cochrane risk of bias tool is probably the most well known and the most used. However, it is not perfect. See the following article.Article Evaluation of the Cochrane tool for assessing risk of bias i...
Thanks Dr. Warren, appreciate your help. Will dig into the article.
In your experience, is it enough to state risk of bias in a qualitative way (high-medium-low) or do I need to specify it in a quantitave basis?
Thanks again,
Cochrane's tool uses labels of high, low, and unclear and that seems to work for me. When doing the meta-analysis, we'll do a sensitivity analysis to see how the results are affected if the studies with a high risk of bias are excluded.
Thanks again, Dr. Warren; it is my first systematic review (SR) and I was kinda lost. In this case we are not doing a meta-analysis just a SR.
Will discuss this insights with the other authors!!
A systematic review which does not include meta-analytic studies (for example, using case series or case reports) is generally regarded as low quality, with an inherent high risk of bias. The Cochrane tool is indeed useful but more relevant to meta-analysis. I guess you just have to state that bias exists, and whether you have been able to take any steps to minimize it.
Dear Dr. Das Poduval,
Your reply was really useful. I was struggling to understand that point. Indeed our systematic review only includes case-series and case-reports, we will state the high risk of bias in our discussion. Do you have any article/paper to support that statement? (so we can reference accordingly).
Really appreciate your help!
Best,
Glad to help. You could go through our paper titled 'Reliability of FNAC as a single diagnostic modality for nodular neoplastic lesions of the thyroid- a systematic review'. The full text is available on Springer for a fee. Even I don't have the full text of the published manuscript because of restrictions by Springer, but if you wish I could send you the section of the raw draft which discusses the issue of bias in the article.
Thanks for your information. Just in case, I have attached your manuscript, you can save it for your records :) all thanks to sci-hub!!!
Will read it carefully,
Have a good night
- Badges
- Science topic