Dear Ihsan, 

Systematic review is integrating different research findings of a single issue. Meta-analysis is using quantitative statistics to integrate different  findings. thus to pool the quantitative  results together i.e. to do meta-analysis, results of the outcomes should be reported by similar study units  or variable categorization.  As most journals and experts say, there should be at least 2 studies to pool the results together (to do meta-analysis).

However,  there is systematic review  including only a single study published in Cochrane databases of systematic reviews. you can see it "Omeje I, Okwundu CI. Effectiveness and safety of first-line tenofovir + emtricitabine + efavirenz for patients with HIV. Cochrane Database of Systematic Reviews 2012, Issue 2. Art. No.: CD007276. DOI: 10.1002/14651858.CD007276.pub2" 

You can access a systematic review  including three studies on the following address: " https://www.hindawi.com/journals/art/2017/5792925/"

meta-analysis is a statistical technique which is not a requirement in a systematic review and should only be used when justified and appropriate. Having or not having a meta-analysis does not affect the quality, reliability or validity of a systematic review (see http://handbook.cochrane.org/ chapter 9 for more information). There are several good examples where despite having many studies no meta-analysis were done in a systematic review because the studies were heterogeneous.

Hi,

Technically to do meta-analysis it is nescessary to have at least 2 studies corresponding to the inclussion and exclusion criteria. Practically, as much as possible/all that corresponding to the inclussion and exclusion criteria and quality standard should be included.

Thanks Mr.Ihsan for raising this question. I am now in the process of doing a systematic review with/out Meta analysis, and this was my inquiry as well, of how many studies to be included?

I do believe that 2 studies should be minimum for meta-analysis in order to  pool the results together. But what If no studies match the same targeted population exactly, and the studies design are different do i still can add them and pool them all together? 

I think that if 2 or more than 2 studies are not matching, it can become a systematic review only.

Dear Nora, 

To pool results together it is not necessity to have similar population and same study designs. the most important element is having similar outcome defined in the same way. for instance if you want to do meta-analysis on HIV about CD4 count, this outcome may be defined as =400 cells per mm3 or as continues variable without categorization or may be classified as =50 cells/mm3.  but the design and the population in these studies may be different. if you used =400 cells per mm3 classification, you need to integrate articles defined CD4 as =400 cells per mm3. or if you used  =50 cells/mm3 you need to integrate articles defined CD4 as =50 cells/mm3 . it is not correct combining

Thanks Dr. Legese, for the great example and explanations. clear enough now !