The Question Under Every Personal Care Hours Fight: What Is the Waiver Actually Buying?
When a state or plan reduces someone's personal care hours, the argument that follows is almost never about the person. It's about a number a tool produced — and most of the people arguing can't say what the number measures. Six things we've learned building an hours instrument that has to survive a fair hearing.
1. Hours are two different quantities wearing one unit. Part of a functional limitation converts to transferred labor — minutes of someone else's hands doing a task. Part converts to presence — a person who must be there for the episode, whatever their hands are doing. Labor is a gradient across dependency levels; presence is a step that saturates early. Tools that only count task minutes are wrong by 5–7× on exactly the cases that end up in hearings.
2. Most assessment forms cannot even express the difference. Between "independent" and "needs someone present" there is one rung — setup help only — and four of the five major state instruments can't record it. On those forms, presence is assumed for anyone not independent, never assessed. That's not a scoring choice; it's a structural blind spot.
3. The case law has a shape, and it's consistent. Courts have struck automatic percentage cuts for living with a caregiver (presumption instead of assessment), hours derived from historical utilization instead of assessed need, and reductions with notices that don't show the arithmetic. The judicially blessed shape: individualized, transparent standards, disclosed variables, and an appeal that can reach the calculation itself.
4. "Informal support" requires three findings, not one. Federal rules define natural supports as provided voluntarily. Available is not willing; willing is not sustainable. An offset that skips from "a daughter lives there" to "minus two hours a day" has skipped the two findings that decide the case.
5. A minute table without a deviation valve is a rigid cap. Every defensible system has a documented route above and below the range — and some states put it in writing. New York's guidance forbids task tools from setting rigid time caps at all. If your table has no exception path, the table is the policy, and that's the problem.
6. The notice is the product. A reduction the member can't reconstruct — which inputs, which minutes, which offset, what changed since last time — fails before any argument about adequacy starts. If the tool changed and the person didn't, the notice has to say so.
We built all of this into a workbench: build a standard from published time-and-motion values, test one member end to end, run the legal integrity flags (each one cited to the case or regulation that decided it), and generate the notice that shows its arithmetic. It runs entirely in the browser; no member data ever leaves the machine.
The thread through it all: an hours number you can't defend line-by-line isn't a determination — it's an opening bid. And the people on the other side of it can't afford to negotiate.
If your organization uses an assessment tool for hours: could you show a hearing officer, today, exactly how one person's number was computed?