Two changes that I am considering for D&D 5th Ed.:
Thinking of hit points as the ability to shrug of scrapes and bruises until your attention and reflexes are suffering enough for one unlucky hit to slip through your defenses has long seemed to me as the most sensible way to imagine damage in D&D. The spell name cure wounds does imply otherwise, but it’s still the most plausible reasoning for why being down 40 hit points with only 1 hit point left will have no negative impacts and you can be as good as new after a night’s sleep. If you can fight, run, jump, and lift without any penalty, then you’re not really injured.
However, when you do get from 1 hp to 0 hp, you’re in immediate danger of bleeding out within literally seconds. That clearly is a potentially fatal injury. That a bandage and an hour of rest can get you back to full strength just can’t be reasonably explained when you are playing a campaign that puts the “fiction first”. It’s an issue that simply can be ignored completely without causing any problems. But it can’t make sense.
Personally, I quite enjoy it in action and adventure fiction when characters have to adjust or completely change their whole plans because of an injured ally they don’t want to risk dying. It’s a great narrative complication, and one that I think can add a lot to a campaign. My goal here really isn’t to be any more realistic, but to add complications and dramatic tension to the game.
The new rule for wounds is pretty simple. I think that’s actually always an important part of any new house rule. You want the players to have to learn and remember as few new things as possible. Whenever a character drops to 0 hit points, he suffers 1d4 levels of exhaustion. This means a character who already has 2 levels of exhaustion has a 25% chance to die immediately. Since exhaustion is not that common I consider this an acceptible increase of risk. And it does provide a strong incentive for exhausted/wounded characters to avoid combat.
The DMG suggests an alternative rule for handling rests that increases the time for a short rest to 8 hours and for a long rest to 7 days. On the one hand, I favor adventures that have few fights often separated by days, so I like the idea of the PCs not being at full strength at the start of every day. Giving them only a short rest every night has some appeal. But when you have to rest 7 days for a long rest, then you really need to retreat to the safety of a town (and one that isn’t currently under attack). Effectively, this means that long resta only happen between adventures. And I like adventures that have the characters in the wilderness for a week or two. No long rests during a journey into the wilds seems way too extreme.
The main things that a long rest provides for most characters is the restoration of all hit points, the regaining of Hit Dice, and perhaps most importantly the regaining of spell slots, amd the ability to switch out prepared spells. Hit points can also be regained through spells, so spell slots really is the big problem when you can’t take long rests during adventures. However, both wizards and druids have the class feature of Recovery, which allows them to regain some spell slots during a short rest, once pet lomg rest. I think this makes for a good mechanism to regain some spells for all spellcasters. When you finish a short rest, you recover expended spell slots that have a combined level that is lower or equal to half your level. Unlike the class feature, this ability can be used on every short rest. If you cast very few spells over a couple of days, you can recover all of your spell slots.
I think the biggest uncertainty with this variants are warlocks. With their really heavily limited number of spell slots, this class is perhaps the most dependant on short rests. But on the other hand, this isn’t going to make a huge difference when there are only two or three encounters in a day, as they also have several invocations that don’t require spell slots at all.
One important thing to keep in mind is that this change primarily affects “narrative time”. Whether the GM says that you rested for 8 hours or for 7 days, does not greatly impact the actual “tactical time” of encounters. Not letting the party rest for a week in the wilderness is the same as not letting it rest in a dungeon for 8 hours. It’s just fluffed in a different way. However, a GM would also have to keep this in mind when considering the frequency of random encounters and the scale of the wilderness. It’s not quite the same thing, but I don’t think it will be too drastically different if this is kept in mind.