When a Special Needs Trust Is Not Enough
Disclosure: I am the founder of Anya, an AI tool built for the problem this article describes, and I have a financial interest in it. I built it because parents need something that works.
Half of caring for a loved one with lifelong support needs is what you do: medications, schedules, paperwork. The other half is the thousands of things you just know: what a sound means, what worked last time. The first half can be taught. The second has to be lived, and what it teaches has never had anywhere to be kept.
That second half has no line in a care plan and no clause in a trust. It lives with the parent, and some of it is lost each time someone else has to step in.
The half you do and the half you just know
The first half of caregiving is direct care: the medications, the appointments, the forms, the schedule, the hands-on work of the day. It is skilled, demanding work, and the field has built a great deal to support it. Much of it can be written down, trained and handed off. The facts of the day can be learned quickly: what time breakfast is, which medication comes after it, and where the day program is.
The second half is different. It is the exact tone that settles your daughter. The three-second pause that means I need a minute, not I'm ignoring you. The difference between the sound your son makes when he's frustrated and the nearly identical sound he makes when something hurts. The routine that looks unnecessary to someone new but took years to get right.
No one sat a parent down and taught any of that. It was learned by being there, across ordinary mornings and difficult afternoons, in moments that would never make it into anyone's notes.
Parents have a phrase for it: I just know. Researchers have heard the same words. A synthesis of studies on parents of people with profound intellectual and multiple disabilities found that parents describe their knowledge of their son or daughter as intuition, a gut feeling, a sixth sense, built through years of direct contact rather than instruction (Kruithof, Willems, van Etten-Jamaludin, & Olsman, 2020). Much of this research studies parents of people with the most significant support needs. The pattern it describes is one parents across the wider range of lifelong support needs will recognize, even where the research has not yet been done on their situation.
It can sound like instinct, but it is what years of paying attention leave behind. That is the other half of caregiving: what a parent just knows.
Why "I just know" can't simply be told
In 1966, the philosopher and chemist Michael Polanyi wrote a sentence that people who study expertise still quote: "we can know more than we can tell" (Polanyi, 1966). Someone who rides a bicycle well can't fully explain how she stays upright.
Polanyi drew a distinction that explains the other half. There is knowing: the live act of reading a person, in the moment, which exists only in whoever is doing it. And there is knowledge: what that knowing leaves behind when some of it is put into words. A note. An observed pattern. "When she goes quiet at dinner, check whether something changed earlier in the day."
Knowing comes first, and knowledge is what it leaves behind. The knowledge can be kept. The knowing can't be handed to anyone; every caregiver has to build their own.
Research on caregivers describes how that knowing is built: through "a constant process of action and response that transforms those involved" (Hoogsteyns et al., 2023). In a later study, parents were asked to video-record the subtle behaviors they expected other caregivers to miss or misread. One mother described what happens when that knowledge goes into a file: the nuance "is almost always lost in that writing" (Kruithof et al., 2024).
The difference matters more here than in most places. A note that says "calm her with music" is right in some moments and wrong in others, and someone has to judge, on the spot, whether this is the moment the note was written for. And most of us can correct a wrong read; when a new doctor gets our history wrong, we can say so. A person who can't fully tell their own story often can't make that correction in a way a stranger would recognize, so a misread gets acted on instead of fixed. It is usually someone who knows them well who notices. That doesn't put anyone's reading ahead of the person's own voice. Whatever they can tell or show about themselves comes first.
What the field built, and which half it was built for
None of this is for lack of planning. Parents in this world are world-class note takers. And reluctant list-makers. Around them, families and the systems that serve them have built tools in good faith: service plans, medical records, medication lists, person-centered plans, health passports, care binders, letters of intent, communication profiles, special needs trusts, guardianships and conservatorships.
Each exists for a good reason. A medication list tells someone what to give. A person-centered plan organizes services around the actual person rather than a diagnosis, a shift that has done lasting good. A health passport puts the essentials in front of a hospital team at the moment a family may not be in the room. A special needs trust protects assets and eligibility for public benefits. A guardianship or conservatorship, or a less restrictive arrangement such as supported decision-making, gives someone standing to act.
Most of what parents are handed is built to hold what can be written down ahead of time: diagnoses, medications, schedules, contacts. Those tools do that job. Some, like communication profiles and person-centered plans, try to hold more. None of them was built to keep what a parent just knows current and usable across a lifetime of handoffs.
I saw what that means at a trust signing. The parents had come to me to plan for their sixteen-year-old son with autism. When every document was signed, notarized and scanned, I handed them the binder of originals, with a flash drive as a backup, and said, "Excellent plans. This is a big day. Congratulations."
And then the tears came. "This is all great legal work," they said, "but how will they know our son?" Since his diagnosis at age four, they had dedicated themselves to helping him: thousands of hours, years of work. "But now what happens if we are gone? How will others know what to do? How can we stop him from regressing without us here?"
Each one is strongest at the question what to do. None of them can keep up with how to know: how to read this particular person's discomfort before it becomes a crisis, or how to tell fear from fatigue from refusal, and when a no is a decision to respect, for this person, specifically. The letter of intent, the most familiar of these tools, gets its own treatment in our article on the Letter of Intent.
The health-services literature has a related idea, informational continuity: passing information between providers so that care isn't fragmented (Haggerty et al., 2003). It is built around a hub, usually the parent, relaying information forward while she is still there to do it. The same literature names relational continuity, an ongoing relationship with the same caregiver. Continuity of knowing is what happens to that understanding when the relationship ends. A small study of family caregivers, in a different population and by a different research team, found the same weight landing on parents: bringing each new caregiver and professional up to speed fell to them, and one parent called it "crazy exhausting" (Tennant, Allana, Mercer, & Burns, 2022).
Even a structured, facilitated group conversation between parents and professional caregivers about the future, evaluated directly, left fewer than half of participants reporting a concrete intention to follow through (Kruithof, Olsman, Willems, Volkers, Kleijwegt, & Nieuwenhuijse, 2024). A conversation, however good, happens once.
Continuity of care and continuity of knowing are not the same thing
Imagine a new caregiver arrives on Monday morning. The plan is there. The medication list is current. The schedule is printed. On paper, continuity has been achieved.
But the parent may still spend Sunday night thinking about all that this person doesn't know. That her daughter saying "I'm fine" can mean the opposite when she says it too quickly. That her son's refusal to get into the van sometimes has nothing to do with the van. That the routine everyone thinks is about preference is actually what makes the transition possible.
These aren't missing instructions. They are pieces of an understanding, and the understanding was not created when somebody wrote a plan. It accumulated through years of living alongside one particular person.
One family I worked with saw this when their son, at forty-five, moved out of the family home for the first time. The move went well, and he settled in. About two weeks later, the staff told his parents he wasn't responding well to their guidance and instructions. "It's hard to get him to pay attention." His parents thought about it, and then it hit them. He wears hearing aids, and to be heard, you have to speak to him face to face; otherwise he filters out the other voices and sounds around him. "We forgot to tell them that." It wasn't attention. He couldn't hear them.
A move-in profile could have held that detail, if anyone had thought to write it down. That is the point. A checklist written before a move is expected to anticipate every detail, big and small, and parents don't carry what they know as a folder they can search on demand. It comes back to them when a situation calls for it. This one came back two weeks late.
Continuity of care asks whether support continues when caregivers change.
Continuity of knowing asks whether the understanding of the person, built through years of lived experience, continues too.
That doesn't mean the parent's knowing passes to someone else. It can't. It means each next caregiver builds their own knowing from what has been kept, instead of from nothing.
The first can be organized through plans, records, funding, legal structures and trained people. The second is harder, because what a parent knows isn't sitting in one place waiting to be transferred. Some of it is in the medical record. Some is in a binder. Some is in text messages and in stories told to family. Much of it has never been recorded anywhere at all.
The problem is not that professionals or future caregivers don't care. The problem is that they haven't had the years the parent had. And each of them, given those years, learns things the parent never saw, which belong in what is kept too.
What happens to the other half: the Knowledge Vulnerability Gap
The loss happens in pieces, because care changes hands all the time: a new direct support professional starts, a longtime caregiver leaves, a sibling becomes more involved, a day program changes, a person moves, a parent gets sick, someone unfamiliar steps in for an afternoon. Each change asks the same question on a smaller scale: how much does the next person have to learn again?
The knowledge erodes between handoffs too. What is told gets simplified with each retelling. What is written goes stale as the person changes.
There is a name for the risk this creates. The Knowledge Vulnerability Gap is the risk that grows for a person with lifelong support needs as what their parent knows about them, both what to do and what the parent just knows, is lost over time and with every change of caregiver, until care runs only on what can be written into a schedule and a procedure. It grows not from anyone's failure, but because there has never been a practicable way to keep that knowledge.
Parents already pay for the gap, a little at a time: the first weeks with someone new spent learning by trial and error what the parent already knew, a misread cue that could have been avoided, an hour spent walking someone through what took years to learn. Physicians who specialize in this population report that, with staffing shortages and frequent personnel changes, professional caregivers may never become expert in the person they support (Huisman et al., 2026). Parents interviewed about being outlived by their son or daughter described fearing the loss of three roles at once: expert, advocate and attentive presence (Kruithof, Olsman, Nieuwenhuijse, & Willems, 2022). That fear is the subject of "I'm Not Allowed to Die First."
For most people with lifelong support needs, there will not be one handoff. There will be a lifetime of them: a parent, then perhaps a sibling, then other trusted next caregivers, with paid supporters coming and going throughout. In my work with families across every stage of life, I have seen the number of caregivers over one person's lifetime run to many dozens, and sometimes past a hundred, depending on the level of care and the specialties involved. The gap grows with each one.
Why nothing has held it
Part of the reason nothing has closed the gap is that three things keep changing at once: the person, the caregiver, and the circumstances around them, all moving through time. A record written once describes a single moment of all three. It starts going out of date the day it is finished.
The rest is in what has been tried. Between August and September 2026, we ran five verification passes, searching published research, California's service-category definitions, and the products and templates sold to families, for any model, legal, institutional or commercial, that keeps this knowledge continuously across caregivers. We didn't find one. What we did find falls, by our classification, into two groups, and each tends to fail in its own way.
Static by design. The letter of intent, the planning workbook, the health passport. These freeze when they are finished, and they tend to fail because they stop being true.
Flexible but unprompted. A notebook, a shared document, a general-purpose note-taking app, with or without a template to start from. These never freeze, but they put the whole ongoing burden of noticing, deciding what matters and writing it down on a parent with no spare hours, indefinitely, on top of the caregiving itself. They tend to fail too, by abandonment. A template settles which categories to use; it doesn't settle who keeps the thing up for forty years.
Underneath both failures is arithmetic. Keeping this knowledge for even one family has always meant skilled, repeated work: drawing out what a parent knows, asking the follow-up questions that reveal why a pattern matters, organizing scattered stories into something usable, and doing it all again every time something changes, for as long as the person needs support. Few families could pay for that, and few service systems were ever set up to provide it.
What changed
There has never been a practicable way for a family to keep this knowledge across a lifetime of caregivers. Now one can be built.
Generative AI does not change what Polanyi identified. Knowing still cannot be transferred, at any level of technology. What changes is the cost of the work around it. A system can now be designed so that a parent records what she learns in fragments, in her own words, as things happen, by speaking rather than writing. Each entry can carry its date and its circumstances, including what didn't work. A conversational system can be designed to ask the follow-up question a busy interviewer would skip, to keep what was observed separate from what someone thinks it means, and, years later, to let a different caregiver find what's relevant by asking about what they are seeing, in their own words, even when they aren't the words the parent used.
None of that gives the next caregiver the parent's knowing. It changes where the next caregiver starts. They still build their own knowing through their own presence. But they can begin with the baseline, the history, what was tried and the circumstances each lesson holds under, instead of with a schedule and a medication list.
The gap is not speculative, and neither is the pattern in what has been tried. What a system like this achieves is a hypothesis, and it is really two. The first is that a parent can keep the record current, in fragments, with far less effort than any binder ever asked of her. The second is that a caregiver who starts from that record reaches familiarity with a person sooner, with fewer misreadings, than one who starts from a plan written once. Both can be tested. If either doesn't hold, the claim is wrong.
Being buildable doesn't make it automatic. Anything trusted with this has to show where each answer came from and keep working after the parent is gone. It has to be one place, held by the parent and then by whoever holds it next, kept current across every age and change of circumstance, because what a parent knows is not fixed. It changes as the person changes, and the record has to change with it. And the next caregiver shouldn't have to study it. They should be able to just ask, see where each answer came from, and still come to know the person themselves. None of this replaces paying and keeping good caregivers, or the time they need to come to know someone. It changes what they start from.
That knowledge is kept by the parent, and then by whoever keeps it next, for the person it is about. It does not belong to any company that builds the tool. The person's own voice is part of it, as it has been all along in a parent's care. For a son or daughter who can tell their own story, or take part in telling it, there is finally a place for that story to grow that doesn't depend on a parent or a caregiver remembering it.
Where a parent can start
Legal and financial planning runs on its own clock, and nothing here suggests delaying it. A trust, a will and the right legal authority belong in place whenever a family is ready to put them there.
The knowledge runs on a different clock. Putting off the day a parent can no longer be there is one thing, and it can wait for as long as it needs to. Putting off the knowledge is another. It changes as the person changes, and it doesn't wait.
So it doesn't have to begin with planning a distant day. It can begin with something that happened this morning. A routine that turned out to matter. A cue someone else missed. Something that worked. Something that didn't. A sentence you found yourself saying to a new caregiver for what felt like the hundredth time. Write down, or say out loud, what happened, what you tried and what the circumstances were, with the date.
Whatever your son or daughter can tell or show you about whether you read them right, a word, a sound, a look you know, belongs in the record too. Their own voice comes first, ahead of anyone's reading of it.
One thing you know today. Then another.
Thirty years at the table
For thirty years, I worked as a lawyer with parents planning for their sons and daughters with intellectual and developmental disabilities, at every age and stage of life.
Over time, a pattern emerged. Parents would come to me with their estate planning already done: the will or living trust completed, the special needs trust drafted, the people with legal authority named, the documents signed and placed neatly in a binder. I would review it all and usually tell them it was in order: legally sufficient as a plan for passing on what they owned and protecting their son's or daughter's eligibility for public benefits.
And again and again, those parents would answer: "Yes, but that's not enough. We need more."
They were right. The "more" was a lifetime of knowing their son or daughter.
Not in general terms. In particulars.
What worked. What didn't. What had already been tried. When an apparent problem was really a sign of something else. Why a small routine mattered so much. What brought joy. What caused fear. How their loved one communicated when words weren't enough. What a new caregiver was likely to misread.
All of that shaped the care their son or daughter received every day. Little of it was in the binder, and nothing I could draft would carry it to the next person.
Those parents were pointing at two planning problems sitting side by side. The first was one that lawyers, financial planners, service systems, and care professionals already knew how to work on: How do we make sure care continues?
The second was quieter, and no less consequential: How do we make sure the knowing continues?
The first had an infrastructure. The second lived mostly in the parent's head.
The question underneath future planning
Families should keep making legal plans, financial plans, person-centered plans and health passports. Continuity of knowing replaces none of them. It names something they were never designed to keep current across a lifetime of handoffs.
For decades, when parents asked what would happen to their son or daughter once they could no longer hold everything together, the field concentrated, understandably, on whether the structures of care would survive. That work remains essential.
But there is another kind of continuity worth keeping: the accumulated understanding of one human being by another. The context. The observations. The corrections. The things that worked once and don't anymore. The things that took years to learn.
The question is no longer only whether someone will be there to provide care. It is whether the next person will have to begin learning your loved one from zero.
Michael Pearce is a former California attorney who spent thirty years in special needs trust, conservatorship and public-benefits planning for families of people with intellectual and developmental disabilities (State Bar No. 145481, inactive). He is the founder of Anya, an AI tool for the continuity of caregiver knowledge built by Tenerra, Inc., and has a financial interest in its success.
Sources
- Haggerty, J. L., Reid, R. J., Freeman, G. K., Starfield, B. H., Adair, C. E., & McKendry, R. (2003). Continuity of care: a multidisciplinary review. BMJ, 327(7425), 1219–1221. https://doi.org/10.1136/bmj.327.7425.1219
- Hoogsteyns, M., Zaal-Schuller, I., Huisman, S., Nieuwenhuijse, A. M., van Etten-Jamaludin, F., Willems, D., & Kruithof, K. (2023). Tacit knowledge in dyads of persons with profound intellectual and multiple disabilities and their caregivers: An interpretative literature study. Journal of Applied Research in Intellectual Disabilities, 36(5), 966–977. https://doi.org/10.1111/jar.13134
- Huisman, S. A., Kruithof, K., Hoogsteyns, M., Nieuwenhuijse, A. M., Willems, D. L., & Zaal-Schuller, I. H. (2026). From hidden insights to better understanding: Physicians' perspectives on caregivers' tacit knowledge. Healthcare, 14(1), 25. https://doi.org/10.3390/healthcare14010025
- Kruithof, K., Willems, D., van Etten-Jamaludin, F., & Olsman, E. (2020). Parents' knowledge of their child with profound intellectual and multiple disabilities: An interpretative synthesis. Journal of Applied Research in Intellectual Disabilities, 33(6), 1141–1150. https://doi.org/10.1111/jar.12740
- Kruithof, K., Olsman, E., Nieuwenhuijse, A. M., & Willems, D. (2022). "I hope I'll outlive him": A qualitative study of parents' concerns about being outlived by their child with profound intellectual and multiple disabilities. Journal of Intellectual & Developmental Disability, 47(2), 107–117. https://doi.org/10.3109/13668250.2021.1920377
- Kruithof, K., Hoogsteyns, M., Zaal-Schuller, I., Huisman, S., Willems, D., & Nieuwenhuijse, A. (2024). Parents' tacit knowledge of their child with profound intellectual and multiple disabilities: A qualitative study. Journal of Intellectual & Developmental Disability, 49(4). https://doi.org/10.3109/13668250.2024.2336084
- Kruithof, K., Olsman, E., Willems, D., Volkers, K., Kleijwegt, B., & Nieuwenhuijse, A. (2024). "What if I'm no longer around?": An evaluative description of a structured group conversation about the care for persons with profound intellectual and multiple disabilities when they outlive their parents. Journal of Applied Research in Intellectual Disabilities, 37(2), e13185. https://doi.org/10.1111/jar.13185
- Polanyi, M. (1966). The Tacit Dimension. Doubleday & Company.
- Tennant, R., Allana, S., Mercer, K., & Burns, C. M. (2022). Exploring the experiences of family caregivers of children with special health care needs to inform the design of digital health systems: Formative qualitative study. JMIR Formative Research, 6(1), e28895. https://doi.org/10.2196/28895