The Best Dementia Care in the World

The data is clear: the best dementia care in the world comes down to two things, specialized person-centered dementia training (not clinical-centered training) and staffing, where the ratio and consistency are the key. At least one caregiver for one resident, through every waking hour, with the same faces doing the caring year after year. Memory care like this does in fact exist, it's rare... but most of what is considered famous, innovative or high-end does not pass this simple test. The reasons are many but they come down to payroll economics and staffing shortages, especially where it's expensive to live.

What does the best dementia care actually look like?

Decades ago a British psychologist named Tom Kitwood studied why two facilities with similar buildings and budgets could produce completely different outcomes for their residents. His answer, published in Dementia Reconsidered: The Person Comes First ^[1], reset the field. Care quality lives in the small interactions, minute to minute, between a person with dementia and the people around them. What was at stake, he argued, was personhood: the standing of being treated as a full person rather than a diagnosis. And he showed that it's maintained or destroyed in those same small interactions. A bath given by a hurried stranger erodes it. A cup of tea made the way she's liked it for fifty years restores it.

That finding has a practical consequence the industry rarely says out loud. If quality lives in interactions, then quality needs exactly two things: caregivers trained to do this work, and enough of them that each resident actually gets the interactions. Training without people is little more than an aspirational poster on the lobby wall, and people without training is just expensive babysitting. The best dementia care in the world is simply the place where both are true at once: person-centered training and at least one caregiver per resident throughout the whole day. Everything else follows from this.

Training needs unpacking, because the word covers two different jobs. Clinical training is what most facilities mean by it: medications, charting, safe transfers, wound care, fall prevention. That training is necessary. It's also a poor measure of dementia care, because the medical part of a resident's day is small, a few minutes of pills and checks. The rest of her day is ordinary hours: getting dressed, eating breakfast, walking somewhere, being talked with. Person-centered training is the skill set for those ordinary hours. It teaches a caregiver that when a resident insists it's 1974, you let it be 1974: correcting her makes the fear worse, joining her calms it. It teaches how to tell agitation from untreated pain, and how to give a bath to someone who finds bathing frightening. None of this requires a nursing degree, and it's shorter and cheaper to teach than a clinical credential. That's exactly why the industry undervalues it. American facilities hire for the credential, because staffing rules and lawsuits count credentials, so the few medical minutes get covered well while the ordinary hours go to whoever is on shift. The best care flips that: it hires and trains for the ordinary hours, and brings in nursing for the minutes.

Why the ratio is everything

Person-centered care is now the most copied phrase in eldercare marketing, so here's the test that separates the marketing from the reality: divide the number of residents by the number of caregivers actually present with them.

At one caregiver for one resident (1:1), the caregiver knows the agitation today started after the phone call from her son. She knows he walks at ten and naps at two. She's standing there before a hard moment becomes a crisis. At one caregiver for six or eight residents, the daytime ratio at a typical American memory care unit, nobody is standing there. Overnight, American ratios commonly run 1:15 or thinner. At those ratios, behavior that a trained caregiver on the spot would simply absorb becomes a sedation order or a discharge letter.

The same arithmetic explains what happens to the workforce. American direct care work pays a median of about $17 an hour, among the lowest paid occupations in the country, and more than a third of the workforce lives in or near poverty, with nearly half relying on public assistance ^[2]. Caregiver turnover has run between 65 and 77% a year across recent industry benchmarks ^[3], so whatever the training program taught walks out the door annually, and the resident meets a stranger again. Continuity isn't a luxury add-on. For a person whose memory is failing, a face or a voice she recognizes on some level calms her in a way nothing else does.

All dementia care falls into 1 of 4 buckets

Training and ratio are separate things: a facility can have either one without the other. That means every dementia care arrangement on earth lands in one of four buckets:

• Trained caregivers at 1:1 or better. The mechanism works. This is the ideal, the best, and it's rare because of what a 1:1 payroll costs and how scarce care workers are where living is expensive. Not because anyone lacks the philosophy or the will, and not even because families lack the money: in most of the West, no one sells it.

• Trained caregivers spread too thin. They're taught the method, then given too little time to use it. This is the best case in most institutional care, including famous European models like De Hogeweyk, the Dutch dementia village, and the copies it inspired in France and Canada.

• A devoted person, untrained. The loved one doing 1:1 care at home. They are there every waking hour, but teams, training and methodology are missing, and there's no shift change, ever. This situation usually ends with the caregiver's own collapse of some kind.

• Neither. The standard product at most facilities: rotating minimal staff, burnout, no dementia-specific training, ratios set by what the labor market will bear.

With those four buckets in mind you can evaluate any facility in the world in a few minutes.

What about the dementia villages?

The most famous dementia care in the world is a Dutch village. De Hogeweyk, opened outside Amsterdam in 2009, looks like an ordinary neighborhood: a supermarket, a theater, a pub, gardens, and around 150 residents living in shared houses while some 240 staff in street clothes keep the town running ^[4]. It deserves its fame for one big thing. It proves to the whole world that people with severe dementia don't belong in hospital corridors, and that they do better living something that looks like ordinary life.

Hold the village up against the four buckets, though, and it lands in the second one. The shared houses run six or seven residents per caregiver. Residents are sorted into lifestyle categories rather than known one at a time. The town itself is a set: the shop and the square are staged so that life looks normal from a distance, even where the staffing can't make it normal up close. Sixteen years on, in a country with some of the best health services research anywhere, reviewers still find no controlled studies showing the model changes outcomes ^[5]. The village is the best version of institutional care, and that's a different thing from the best care.

The places that copied the village say the same thing in numbers. France's village was built with over 28 million euros of mostly public money; residents pay about 2,000 euros a month and the rest is subsidized (for citizens) by the government ^[6]. Canada's privately paid version in British Columbia costs an estimated $70,000 to $90,000 per resident per year ^[7]. The American adaptation is a day program in California, a staged 1950s main street you can visit for about $95 per eight hours ^[8], because the arithmetic of a residential version doesn't work at American wages and living expenses. Wherever caregiver hours cost more, the design budget stands in for the staffing budget: more scenery, fewer people. Every prop, in the end, marks a missing person: the pretend bus stop exists because no one has the time to ride a real bus with them.

Where 1:1 care actually exists

So the search for the best dementia care in the world turns into a sharper question: where can a facility afford to hire, train, and keep one caregiver for every resident, and pay them well by local standards?

Almost nowhere in the Western world. A 1:1 trained care facility in the US would cost around $26,000 to $32,000 per resident per month before profit, which is why no US chain offers it at any price. The $8,000 to $15,000 a month memory care that American families pay for only buys the 1:6 to 1:8 ratio ^[9]: not good, and certainly not the best in the world.

The Best in the World model only works where the cost of living is low but the care profession is strong. In Thailand, employing one caregiver for a year, wages, training, and benefits included, costs a fraction of what it does in America. Care work is also a respected, high-status profession in Thailand, fed by universities and nursing schools and paid well above the local cost of living, which is why caregiver turnover runs 20 to 30% as compared to the US average of 70% per year. Real estate, food, transportation and energy are much cheaper too, by the same proportion. There's also no placement broker industry taking commissions and no real estate investment trust raising the rent every year, so nothing in the price adds to the cost without adding to the care. The result isn't cheaper mediocre care, it's affordable extraordinary care. It's the only cost structure anywhere in which one trained caregiver per resident, the same faces year after year, and doors open to a garden every single day come together as a product an ordinary family can buy. The best dementia care in the world costs about $3,500 a month, covering housing, meals, nursing, and the care itself, and it can't be rebuilt in Cleveland or Los Angeles or Florida at any price.

That's also why, when one published country comparison weighted eleven factors from staffing to training to climate, Thailand scored 90.55 of 100 and the United States 36.15, last among the countries ranked ^[10].

The two-question test

The best dementia care in the world comes down to two things, so two questions measure any facility on earth:

1. What's your caregiver-to-resident ratio? (You want 1:1 or better.)

2. What are your caregivers taught about dementia beyond the clinical basics, and what is their turnover percent? For example: What would they say if a resident said it was 1974? (You want to hear a non-clinical focus, and some version of: assume the frame of it being 1974, as opposed to correcting or deflecting them.)

A facility that passes both test questions is the real thing. If you want a tiebreaker, read the discharge clause in the residency contract: American chains commonly reserve the right to discharge a resident whose behavior exceeds what the staffing absorbs ^[11], and the best facilities will see every "behavior challenge" as an opportunity for deeper training.

The best memory care in the world...

is in fact a very real and a very measurable thing: a trained human being, present, known, and still there next year and the year after. Everything else is marketing.

References

1. Kitwood, T. Dementia Reconsidered: The Person Comes First. Open University Press, 1997.

2. PHI. Direct Care Workers in the United States: Key Facts. 2025 report, 2024 wage data.

3. Home Care Association of America. Caregiver turnover benchmarking. 65.2% (2021); subsequent benchmarks ranged to 77%.

4. De Hogeweyk. Official site.

5. CDA-AMC (Canada's Drug Agency). Dementia Villages: Innovative Residential Care for People With Dementia.

6. Center for Cognitive Health. The Village Landais Alzheimer.

7. Langley Advance Times. Langley dementia village cost per patient estimated at $70,000 to $90,000 annually.

8. George G. Glenner Alzheimer's Family Centers. Town Square.

9. SeniorLiving.org. Average Memory Care Costs. 2026.

10. Memory Care Guide. Global Memory Care Quality Index 2025.

11. National Consumer Voice for Quality Long-Term Care. Involuntary Transfer or Discharge.