Ask a treatment program what a course of care costs and you will get a precise answer. Ask what it costs the family to be there and the room goes quiet. That is not an oversight, it is structural: the money never passes through anything you operate, so there is nothing to report on.
Every other cost of care is visible because it moves through a system you run. Lodging does not. The family finds a hotel, pays the hotel, and the transaction is complete without touching your practice management system, your finance team or your patient record.
The consequence is that the largest out-of-pocket cost of a course of treatment, for a family travelling in, is the one cost nobody has a figure for. Not a rough figure: no figure at all.
Three reasons, in ascending order of how much they tend to move a decision.
First, it is a real barrier. Financial pressure around a course of treatment shows up as shortened stays, skipped follow-ups and patients who do not come back, and lodging is a meaningful share of that pressure for anyone travelling in. Second, it is increasingly asked about, by boards, by funders, and by anybody assessing whether a program serves the population it claims to. Third, you cannot improve or defend a programme you have no numbers for.
You do not need a research project. Four fields per visit, captured at scheduling, gets you most of the value.
The first, second and fourth are already somewhere in your systems or easy to add. The third is the hard one, because the payment happens between the family and a hotel.
There are three practical routes. You can ask, on a form at the end of a course of treatment, which gets you a self-reported sample rather than a census. You can estimate, using nights times a market rate for your area, which is defensible for a board slide and useless for anything specific. Or the booking can happen somewhere you can see it, which is the only route that produces an actual number for every patient rather than an estimate for the average one.
The first cut worth looking at is spend against distance band. It usually shows that a small group of patients travelling furthest carries a strikingly large share of the total burden, and that group is normally also the one with the worst completion rates.
The second is spend against whether treatment was completed. If those two correlate in your data, you have stopped talking about a nice thing to do for families and started talking about a clinical outcome, which is a different conversation with a different audience.
A survey gets you a self-reported sample, which is worth having and is much better than nothing. What it will not give you is a per-patient figure you can cut by site, distance or treatment type, or read against completion. For that the number has to be captured rather than recalled.
Because it affects whether your patients complete treatment, and because you are the only party positioned to see it. The family sees their own cost, the hotel sees a booking, and nobody sees the pattern across a population except the program they all travel to.
Add the origin postcode and the number of nights to whatever you already capture at scheduling. Those two alone, read against completion, will tell you whether this is worth pursuing further at your program before you invest in measuring spend.
Some will, if you ask at the right moment and explain why. Response rates on end-of-treatment questions are usually better than people expect. The limitation is not honesty, it is that you get a biased sample: the families who finished.