For that matter, malpractice suits seem like a uniquely US phenomenon.

They still happen here in Scandinavia, but they're fairly rare.

The weird thing is most doctors are not trained to be scientists, but many of them are expected to do science.

There are ways to punish "malpractice" that aren't lawsuits, on many different levels, so I think the status quo argument still holds.

I would guess that, prior to malpractice suits, the main threat to a doctor was a bad reputation, spread by word of mouth among patients. This still provides an incentive to do nothing too weird, but now "weird" means "weird as judged by patients and their loved ones."

Would this have made for more innovation or less? It's not clear to me. On the one hand, patients had less experience in medical matters than doctors, so things that other doctors might have considered normal would have seemed weird to the patients. This might have encouraged doctors to be even more conservative than their peers would allow. Another doctor might say, "Yeah, that's how this rare condition is treated, on those few occasions when it is treated." But the patient hasn't heard of those "few occasions", and so the treatment will just seem weird. If the treatment fails, surviving loved ones might spread rumors about how the doctor killed the patient with weird treatments.

On the other hand, patients would probably often defer to the authority of their doctors. Maybe the patient would think, "Well, this does seem a little weird. But doctors often do things that seem a little weird. I'll trust him if he tells me that it's not too weird." In that case, a doctor might be able to get away with weirder treatments than his peers would allow, because patients, unlike other doctors, wouldn't feel comfortable gainsaying him, even if things didn't work out.