One of the cleverest changes in 4th Edition is the introduction of healing surges.
Previously, healing was flat rate. A healing effect might restore 1d8+1 hit points, and that was its effect regardless of who you used it on. A plate-armoured fighter would get hit less often, and therefore a heal on such a high-defence ally would "last longer", but otherwise healing was more or less indiscriminate.
The problem with flat rate healing is that it penalises high hit point totals. A fighter with 50+ hit points can bankrupt their friends of spells as they desperately try to restore him to full health after every fight.
Healing surges make healing proportional. Using one of your limited supply of healing surges restores a quarter of your HP. You can use as many as you like at any time while not in combat, and you can use them in combat by being targeted by a healing effect or by using your once-per-encounter Second Wind.
Making healing proportional means that healing powers become more tactically interesting. The healing word that gives a wizard 6 HP will provide 10 to a fighter. Clerics and paladins want to be healing high HP classes because it makes their healing more efficient. Conversely, high HP classes want to be taking hits because they're easier to heal.
This promotes the twin 4th Edition design goals of teamwork and tactics. No longer is combat a row of unconnected duals - it now becomes an exercise in attempting to control the overall flow of battle.
It also allows for some interesting sub-mechanics around healing. Clerics, and indeed most healers, heal allies by triggering one of the ally's healing surges. Paladins, however, can use their lay on hands ability to heal an ally using one of the paladin's surges (which they start with an above-average supply of). The management and conservation of surges across the party can therefore become a critical exercise.