Somewhere in a health system this week, a small team is looking at a dashboard. Patient acquisition, referrals, the list of families waiting for a call back. The huddle takes half an hour, happens most weeks, and it is one of the least glamorous things in American health policy. It is also, according to a paper published in the Journal of the American Geriatrics Society, one of the things that decides whether a dementia care program survives its first year. The authors recommend frequent clinical and operational team huddles precisely because problems in these programs show up early, in numbers, before they show up in a family's living room.

The program in question is GUIDE, short for Guiding an Improved Dementia Experience, the Centers for Medicare and Medicaid Services' voluntary nationwide test of a different way to pay for dementia care. Instead of billing per visit, participating organizations receive a monthly per-beneficiary payment to deliver care coordination and care management, caregiver education and support, and round-the-clock access to a support line, with a separate payment for respite so an unpaid caregiver can get a break. CMS announced 390 participating organizations on July 8, 2024. Ninety-six of them, the Established Program Track, began delivering GUIDE services on July 1, 2024, because they already ran most of this care. The other 294, the New Program Track, were given a pre-implementation year and started on July 1, 2025. The model is scheduled to run eight years. Which means the bulk of GUIDE, in numerical terms, has now been operating for roughly one year, and the question of what it actually takes to stand up a comprehensive dementia care program stopped being theoretical some time ago.

That is the gap the new paper aims at. Published online on October 31, 2025 and appearing in the November issue, it comes from the National Dementia Care Collaborative, a coalition of scientific and clinical leaders in evidence-based dementia care, with Kristin Lees Haggerty as lead author and a roster that includes David Reuben of UCLA. The authors draw on six comprehensive dementia care programs that were built and tested before GUIDE existed: Aging Brain Care, Alzheimer's and Dementia Care, BRI Care Consultation, Care Ecosystem, Integrated Memory Care, and MIND at Home. From that collective experience they set out four steps: identify the key leaders and partners who can assess need and feasibility; prepare a value proposition tailored to the local institution; get the program started efficiently; and build for sustainable implementation rather than a demonstration that folds when the grant or the model ends.

Read that list again and notice what it is not. It is not a clinical trial, it does not enroll a single patient, and it reports no outcomes. It is a peer-reviewed implementation paper: expert distillation of what worked in six earlier programs, offered as guidance to health systems inside GUIDE and to organizations that were never eligible for it. The same group has separately described its aim as spreading evidence-based comprehensive dementia care beyond the model's participant list, including to organizations that cannot bill Medicare for it. That is a candid admission of the underlying problem. The payment code is national; the capacity to deliver what the code buys is not.

For a family, the four steps translate into things you can actually observe. A named person who answers the phone and knows your relative's history. A care plan that survives a change of doctor. A respite arrangement that exists in practice and not only on a benefits page. A program that is still there in year three, because someone did the unromantic work of showing a chief financial officer why it should be. The second step, the local value proposition, is the one most likely to determine whether the program near you thrives or quietly shrinks, and it is the one no family will ever see. The gulf between the 96 organizations that were already doing this work and the 294 that had to construct it inside a year is the real story of GUIDE's first phase, and the new paper is essentially a set of instructions written for the second group by people who have been in the first.

What nobody can yet tell you is whether any of it keeps people at home. GUIDE is being tested against exactly that question, along with quality of life, caregiver strain and total Medicare spending, and those answers will come from the model's independent evaluation over years, not from an implementation blueprint published in year two. Until then, the honest summary is this: Medicare has built a payment mechanism for comprehensive dementia care, several hundred organizations have signed up to deliver it, and a group of clinicians who did this work before it had a billing code have written down how they think it should be built.