I am working on a meta-analysis. However, from the sensitivity analysis I got s significant change on the result (but on Egger regression it is okay --> p>0.05).
Illustration:
1. Original finding (11 studies): OR = 0.60 (95% CI = 0.45-2.1, p=0.45)
2. After sensitivity analysis (10 studies --> let's say removing study X): OR = 0.50 (95% CI = 0.25-0.61, p=0.04)
Note: these are not the true value but it resembles our finding
What must I do? Remove study from the final one or still includes original one?
And how to interpret it? Thank you
I think you can not only remove the study which gives different estimates from the others. You need to explore a little bit the characteristics of this study. The sensitivity analysis only gives you a way to see if your results are robust. In this case it is not the case. As recommended by Cochrane handbook (https://handbook-5-1.cochrane.org/chapter_9/9_7_sensitivity_analyses.htm), sensitivity analysis gives you potential ideas for future studies focusing on potential issues. Exploration on why the specific study may differ from the others in terms of experimental setting, outcome measuremement would be the first step to present in your study...
The publication bias tests (like Egger's test) may lack of power to identify the presence of publication bias when the number of studies is limited.
Good luck!
The Egger test and publication bias is largely irrelevant for interpreting sensitivity analyses (unless the sensitivity analysis is based excluding studies with a high risk of publication bias).
How to interpret the results of your sensitivity analysis will depend on why you are excluding this one study. If you have a good reason for excluding this study (e.g. high risk of bias, heterogenous sample or intervention), then you could present the results of this sensitivity analysis, alongside your main results.
For example, '10 studies were at low risk of bias and 1 study was at high risk of bias. When this study was excluded we found a significant effect. Therefore, we conclude this intervention has a significant effect'
Thank you very much for sharing your thoughts, Dr. Sébastien Buczinski Joseph Whittaker.
- Badges
- Science topic
- Similar topics
- Correction