Nearly every program that sees out-of-town patients has a page of hotels, and nearly every one of those pages is older than the person maintaining it thinks. This is not carelessness. A lodging list is the rare piece of infrastructure that gives no signal when it fails, so it fails invisibly for years.
It is worth naming the failure modes rather than treating staleness as a general condition, because each one has a different check.
Once a quarter, pick three properties from your own list and try to book each one for a week about two months out, using only what your page tells a patient. Not by calling your contact at the hotel: by doing what a patient does.
This takes about twenty minutes and it finds broken links, dead codes, closed properties and wrong numbers in one pass. It also tells you something no audit of the page can: whether a patient could actually work out what to do.
Most lists are hard to maintain because of how they are written, not because maintenance is hard. Three changes do most of the work.
The information a travelling patient actually needs is narrower than what most lists provide, and it is rarely the star rating.
Distance and how you get there matters more than anything else, because a patient arriving for a 7am appointment after surgery is not going to walk a mile. Whether there is a kitchen matters enormously for stays longer than a few nights. Whether a companion can stay in the room decides whether the trip is affordable at all.
A list is the right answer for one site with a couple of nearby hotels and somebody who owns it. It stops being the right answer at fairly predictable moments, and it is worth recognising them rather than adding a tenth property.
The usual triggers are a second location, a patient telling you a hotel on the list has closed, or somebody asking a question about your lodging programme that the list cannot answer: how many patients used it, how far they came, what they paid, whether the room was any good.
Fewer than you think. Three to five properties that somebody has actually checked is more useful to a patient, and far more maintainable, than a long list assembled once. Long lists also push the decision back onto a patient who has no basis for making it.
Only if you are prepared to maintain them, which almost nobody is. A stale price is worse than no price: it sets an expectation the patient discovers is wrong at the point of booking, and it makes the whole page look untended.
A named person with a recurring reminder, not a department and not "whoever notices". The single most common cause of a stale list is that its owner changed roles and nothing was handed over, because nothing about the page said it needed handing over.
With a plain list of links, you generally cannot, and that is the deeper problem. Bookings happen between the patient and the hotel, so nothing comes back. If that question matters to you, it needs solving separately from the list itself.