Hello, I'm conducting a meta analysis on a certain marker in diabetics and healthy controls. Some studies divided the cohort into type 1 and type 2 DM and control group. When inputting the data for the analysis do I use 2 entries or do I calculate the average of both DM and enter them as one? for example
A)
Type 1 DM 50 patients - Control 100
Type 2 DM 50 patients - Control 100
B) Average of Type1 and Type2 100 patients - Control 100
The marker I'm using doesn't differ between type 1 and type 2 DM.
You should be consistent when using the result for meta-analysis. In your case, A) should be Type1+2 to be comparable with B). If controls for them ( Type1 and 2) are independent, then combine, If the controls are the same population, then do not. The following may be helpful:
Article Combining Results from Independent Investigations: Meta‐Anal...
Chapter Statistical Methods for Examining Heterogeneity and Combinin...
@shin murakami thanks for the answer. What I meant is that should i go with method A or B? both patient groups have the same control
As far as the controls are identical in both the groups, you should be combining T1DM and T2DM otherwise you will dilute the result by including the identical control groups twice.
Hope this helps.
Dr Asiful
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis