The NHS does approve experimental and innovative treatments, and it does treat rare conditions. It also chooses whether to offer, or not offer, expensive new drugs. This, again, is democratically accountable - if it refuses a drug that is widely seen as life saving, the tabloids will pressure the politicians into turning up the heat and forcing it to change course. Also there absolutely is an incentive to specialize; the NHS pays a doctor's salary, so (provided it's willing to hire them) they do not have to pick a specialty that's profitable.
Moreover, the NHS, being free at the point of delivery, is willing to take on expensive treatment of poor people. In Britain, the idea that someone might go untreated perhaps for years for a quickly treatable illness, because going to hospital would bankrupt them, is seen as abhorrent and barbaric.
Ill health strikes randomly without regard to wealth; health of everybody benefits everybody. Therefore, everybody should pay according to what they can afford, everybody should be covered according to what they need.
I was not targeting the British NHS, in fact, I didn't mention it. However, I would point out that:
> the idea that someone might go untreated perhaps for years for a quickly treatable illness, because going to hospital would bankrupt them, is seen as abhorrent and barbaric.
Is not the same as having a rare condition. You have added (1) quickly treatable, and are assuming (2) easy to diagnose.
Doctors here in Canada can still be sued for malpractice (and with few exceptions, doctors are NOT employees of the government; they are independent contractors), so that entire line of argument is bogus, like every single American argument I've ever heard against a true universal health-care system.
Second, if I were, then small and homogenized could just as easily apply to a geographically concentrated population with the same macro drivers of external illness vectors (climate, weather and epidemiological proximity). Odds are that if the flu is going around the UK, the person coughing has the flu.
[EDIT] Third, I did say might and the short run...sometimes social engineering problems can take a while to diagnose, and can be difficult to treat.[/EDIT]
And just like anywhere that regulations spring up (Canada or not), regulatory compliance becomes a de facto legal defense against malfeasance.
Might work in the short run in a small and homogenized enough population, where everybody basically only needs the same type of care; beware if you ever get a rare condition, since there is no incentive to specialize, and the regulations for "edge cases" are written as they are discovered.
The real solution to high medical prices is to lower costs, foremost of which would be insurance premiums doctors need to pay to cover their practices. This can be accomplished through tort reform. The best part of this is that if you don't like insurance companies taking money, this reduces that purely by market action.