I think you might write to the journal, the author and I rhink the peer reviewer. I would suggest first the joirnal and then the peer reviewer. You might need to explain the causes. It would count as a researcher questioning the article and could say there is more proof,in case there is more data (like a method). Maybe then say you need to speak to the peer reviewer. I am not sure the author receives the notification or maybe (false results or a mistake?). I am not sure at this point the article is then mentioned in previous reports. The final result the article is then either retired or for it to consider the rest, inluding method or procedure. You are right it is best to follow the indication above. I am fairly certain,but I might not said clearly the order. Thank you Anton Vrdoljack for your link.
Well done! you highlighted a blunt unethical issue in the research field, if somebody did this so far he is accountable. Otherwise, nowadays journals have no recruitment with this concern, they just run their businesses to get published, whatsoever are the results? you may run after an author.
The question is how did you discover that the results are fake? If you have a solid evidence, then you can send it to the editor in chief and he can decide what he should do with the paper, if you dont have a solid evidence and somebody else told you that the results are fake, then there is nothing you can do, the editor in chief wont take your word for it.
You have to inform the authors directly and boldly. You have to call a spade, a spade. The particular journal can blacklist the author or authors in their journal in the future. But I am scared that's all you can do!