I'm having trouble choosing which model to use to analyze my findings.

My meta-analysis is about how effective a drug is compared to placebo or no intervention to reducing pain during needle procedures.

So I've collected the studies I need, I'm looking at a specific drug and the outcome has to have a measure of pain. This is all true for all of the studies I have selected.

However, the studies are all different:
in needle procedures: some use vaccinations, others use cannulation, venepuncture. These will have differences in pain.
in population: I'm looking at children only. But this has implications as younger children have a different understanding of pain to older ones.
Application of drug: they apply the drug in different ways across the studies.
Placebo or no intervention: Some use a placebo, some studies use no intervention as a control measure.

And to add: The conclusion of some of the studies do not support the drug while others do!

------- A little info---
A fixed effect meta-analysis model assumes all studies are estimating the same (common) treatment effect. In other words, there is no between study heterogeneity in the true treatment effect. The implication of this model is that the observed treatment effect estimates vary only because of chance differences created from sampling patients.

A random-effects meta-analysis model assumes the observed estimates of treatment effect can vary across studies because of real differences in the treatment effect in each study as well as sampling variability (chance). Thus, even if all studies had an infinitely large sample size, the observed study effects would still vary because of the real differences in treatment effects. Such heterogeneity in treatment effects is caused by differences in study populations (such as age of patients), interventions received (such as dose of drug), follow-up length, and other factors.


So I thought I should use a random effects model. But I've read on other websites that to use a random effects model is to assume that your sample population is RANDOM (so you've taken it from a wider population and its the difference between participants causing the variability) but the studies I'm looking at; they each had a specific inclusion criteria for their participants (children among other stuff) so the difference between the participants is controlled.

So individually, the studies individually are homogenous but comparing across the studies, they are heterogenous in population (some looking at young children, some at older), procedures and other stuff.

Which model do I use? This will affect the interpretation of my results and therefore it is important that I use the right one.

I'm so confused

I hope you guys can help and thank you so much in advance, I appreciate you and your time.