Yes, a fraudulent study should be disregarded and removed from meta-analyses.

But are you trustworthy, or a partisan? Did you call out the meta-analyses that included Surgisphere's paper? Have you called out the ones including the Cavalcanti study despite the admission that 40% of the control group had been recently medicated? Did you call out the Roman meta-analysis for it's pile of mistakes? Did it bother you at all that the WHO tested HCQ using protocols that made no sense to anyone? What about the meta-analyses that include those studies? They look pretty positive after exclusion, or when the inclusion criteria focuses on optimal (early) treatment.

https://roundingtheearth.substack.com/p/how-to-rig-research-the-who-edition-105