As noted, magnetic poles shift around all the time. If you've ever been in a cockpit of an airplane, right next to their magnetic compass is often a "compass correction card" which is periodically adjusted as magnetic north drifts around a bit.
As to the poles completely shifting - it does happen. It's a periodic event that happens every few hundred thousand years. It IS a magnetic shift - not a physical shift. After it completes you'd basically see all compasses pointing south rather than north. The physical orientation of the planet would remain unchanged. The physical angle of inclination of the Earth changes VERY little - less than a few degrees even over billions of years - due mainly to us having a moon that is a significant fraction of the planet's mass. The moon tends to act as a "gravitational anchor" which keeps Earth physically much more stable than a planet like Mars for example, which has two moons but they are very, very small compared to the mass of the planet.
Thing is, though it's something that happens very rarely on human timescales, in geologic timescales it's not all that rare. It's not an extinction event - indeed we don't even see levels of life really dropping at all during pole shift events. IMHO, if it happens anytime soon, it'll be something that really hurts air travel for a while, and we'll need a good solid decade or so to get all of our navigation tech that relies on a compass back into swing, but otherwise - no big deal.
There's also a question of how quickly the change occurs too. If the poles just slowly swap over a period of 20 years or so then I'm sure our current methods of dealing with pole drift will suffice. If it changes in a matter of days - it'll probably be more chaotic. Unfortunately while the geologic record shows pretty clearly that the pole shifts happen, it's rather hard to pull out any meaningful data as to how long the actual change takes.
Worst case scenario as far as health - the Earth's magnetic shield IS what deflects a lot of solar radiation around the planet. If it's disturbed significantly during a shift, we may see an increased rate of cancer soon afterwards, but again - not like an extinction event or anything.