The problem is that studies that show bad results has a lower chance of been written and a lower chance of been published and a lower chance of been collected in the sites. So there is a lot of informal involuntary filtration, that hides the the bad result and collect only the good results.
Someone posted https://ivmmeta.com/#fig_fpr and https://hcqmeta.com/#fcite_rct . It's nice to be able to compare both.
Take a look at this paper: https://c19ivermectin.com/chowdhury.html
One interesting thing is that it compares Ivermectin and Hydroxychloroquine. It is listed as a success in the https://ivmmeta.com/#fig_fpr but it's not listed a failure in the https://hcqmeta.com/#fcite_rct site.
If it had the opposite result, would it be listed in ivmmeta or only in hcqmeta? If it has a null result would any of them had shown it? (Assuming that the doctors took the time to write a boring article about a null result, and the journal published a boring article about a null result.)