A non-zero fraction of what our medicine can do in mice, in fact, works in humans. That is why testing of mice hasn't been abandoned, unlike, say, Galen's theor

A non-zero fraction of what our medicine can do in mice, in fact, works in humans. That is why testing of mice hasn't been abandoned, unlike, say, Galen's theory of bodily humors. Yes, that fraction is lower than you or I would like it to be. Mice are short-lived and cancer-prone, no wonder that the discrepancy between them and humans is big. They even do not get some of the diseases that we do (e.g. the murine models of ALS are pretty useless). But they are cheap, too, which counts. It is possible that, in the future, a lot of the testing will be done with organs on chips, much closer to actual human tissue. But we don't really have anything more efficient right now.