These three get discussed as if they were points on a scale from cheap to expensive. They are not. They differ in what you are committing to, who carries the risk when plans change, and how much of somebody's week it takes to keep working. Pick on those and the price question mostly answers itself.
You agree a rate with one property, usually annually, usually through their sales office. No rooms are held, you owe nothing if nobody books, and the hotel owes you nothing if it sells out.
This is the right shape for most programs starting out, and it is genuinely low-risk. Its weaknesses are all downstream: it covers one property, it needs re-signing, the patient has to invoke it correctly, and it produces no record. See the guide on where corporate codes break for the specifics.
A block reserves a specific number of rooms on specific dates. It solves the availability problem outright, which is the one thing a negotiated rate cannot do. The catch is in the contract: most blocks carry an attrition clause committing you to a percentage of the rooms whether or not anyone sleeps in them, and a cut-off date after which unsold rooms return to the hotel.
That is a real financial commitment made against a forecast of patient travel, and patient travel is the thing that moves. A procedure gets rescheduled, a cohort is smaller than expected, and the difference is an invoice.
The third shape removes the per-property relationship entirely. Rather than negotiating with hotels, you give the patient a way to book from a set of properties chosen around your site, and they book and pay themselves.
What you are trading is a vendor relationship for a maintenance burden. There is no annual renewal per hotel, no attrition exposure, and coverage extends to properties you never negotiated with, including in cities where you have no relationship at all. What you take on is depending on somebody else's inventory and doing the work of choosing that vendor carefully.
The useful question is not which is cheapest, because that varies by property and week and nobody can answer it in the abstract. It is which one your program can still be running properly in two years.
Whichever you choose, the record is a separate decision and it is the one most programs skip. A negotiated rate produces nothing. A block produces a rooming list and an invoice, which tells you about the block rather than about the patients. A platform may or may not, depending on which one.
If you ever expect to be asked how many patients travelled, from how far, for how many nights, and what it cost their families, decide now how you would answer, because none of these mechanisms answers it as a side effect.
A commitment to fill a percentage of the rooms you blocked, with a payment owed on the shortfall if you do not. It is standard in block contracts and it is the reason blocks carry real financial risk for programs whose travel volume moves.
Frequently, and many programs do. A negotiated rate at the hotel across the road plus a broader option for everywhere else is a common and sensible shape. The failure mode to avoid is a page listing both with no indication of which a patient should use.
It genuinely depends on the property, the week and what the public rate is doing at that moment. Any answer that does not start by looking at your specific hotels is a guess, which is why the durable comparison is coverage, maintenance and the record rather than price.
Not if the dates are predictable and you are filling it. Blocks work well for exactly that case. Revisit it when the pattern changes, particularly if you start paying attrition or find yourself blocking rooms for a cohort whose size you cannot forecast.