Another problem with studies is, that negative results are rarely published unless it's something really really "interesting". "we tried treating X with Y, and

Another problem with studies is, that negative results are rarely published unless it's something really really "interesting". "we tried treating X with Y, and it didn't help (even though in theory it should have some effect)" is harder to get published than "we treated X with Z in vitro and it killed all the cancer cells (and noncancer ones too, whoops)".