Redesigning How Home Care Runs
The president of home therapy asked a question the business had avoided: is the way home care is structured actually the most efficient way to deliver it — and can it grow as it stands? A stalled pilot, three live markets, and one improvised measurement system later, the answer reshaped how the operating model itself gets designed.
All figures drawn from the public record; where results remain internal, they are described directionally. Visuals recreated as abstractions — no client artifacts or confidential data appear on this page.
A structure built for yesterday’s scale
Home therapy had grown up one clinic at a time, and its operating model showed it: the same nurse who trained a new patient also managed that patient’s ongoing remote care, on call rotations stretched for weeks — sometimes without rotation at all — and every clinic solved scheduling, training, and care management alone. Leadership was asking the uncomfortable question out loud: can this structure carry the growth the strategy demands?
A first attempt at a new “home service model” — splitting the training function from ongoing care management — was stuck in South Carolina, designed but unable to launch. I took it over while still closing out the New York program, and had it live in under sixty days.
You can’t redesign what you can’t see
The first constraint wasn’t the model — it was measurement. The enterprise scheduling platform was still a future project, and without visibility into how clinicians actually spent their time, no one could ever prove whether a new operating model worked. So we improvised: early enterprise access to Microsoft Bookings, a handful of carefully rationed licenses, designated owners entering every program’s schedule weekly, and a daily synthesis routine — which areas need support, where to level resources, when to inject the float pool. Imperfect, manual, and enough: the pilot was instrumented from day one.
The second discovery was bigger. As South Carolina, Las Vegas, and Arizona ran in parallel, the data kept breaking the assumption underneath the whole exercise — that one model would win. Clinic volume, training pace, and patient acuity changed the right answer market by market. Home therapy runs on the patient’s clock and the physician’s clock, not the company’s — so the honest question isn’t “what is the standard model?” but “how much of this business can be standardized at all?”
Standardize the model — or the method
Standardize the questions, not the answers
The operating model became a design system rather than a mandate: a codified set of building blocks — role architectures, case management formats, scheduling approaches — and a disciplined way of choosing among them. Clinic managers weren’t handed a structure; they were tasked, with their teams, to reason from what they knew about their clinics and markets, drawing on what had worked elsewhere. My piloting rule made the change safe to attempt: I never asked anyone to commit to a new model — only to try it, for a defined period, with honest feedback and a guaranteed way out. Nobody transforms willingly inside a trap.
Two principles were written in as non negotiable. Patients travel last: every reasonable option moves a teammate before any patient is asked to go farther — period. And people are aligned to markets, not buildings: job descriptions were rewritten so clinicians join a market’s care team, ending the accident of geography that had trapped talent behind clinic walls.
A design system — and the people who can use it
- ✓The stalled pilot, launchedSouth Carolina’s home service model live in under sixty days — sub markets, split training and care management functions, centralized regional scheduling.
- ✓Instrumentation before the platformAn interim scheduling and measurement system built from rationed licenses and weekly discipline — proof, in data, that the model worked, and the bridge to the enterprise platform in Case 01.
- ✓A codified model libraryRole architectures — dedicated trainers, float pools, dedicated care management — and four case management formats, organized into a framework leaders could select from and adapt.
- ✓Market attached roles, in writingJob descriptions rewritten from clinic attached to market attached — the structural change that made every other flexibility real.
- ✓Top of license delegation, unlockedA workshop built with the education department on scope, trust, and teamwork — moving care technicians from sidelined administrative work to genuine clinical support, so every role operates at the top of its license.
- ✓Leaders trained to design, not complyScenario based courses embedded in the foundational curriculum for clinic managers, directors, and regional leaders — where the right answer is sound reasoning, not a memorized structure.
- ✓A permanent experimentation hubThe project was never closed. It became the standing engine for operating model innovation — carried market by market across the country as national interest grew.
What the design system returned
The pilots proved out in the numbers. In South Carolina, home dialysis cost per patient per month fell from $619 to $514, held consistently across roughly 900 patients within a year; in Las Vegas, from $800 to $587 across a comparable population — together on the order of $3M in annualized run rate savings, achieved by redesigning the operating model rather than cutting the work. The quieter results mattered as much: on call rotations became humane, clinicians began clocking out on time, and patients — who nurses feared would resist shared care — welcomed having a full team behind them. And the pilots served a second purpose: they were the live design playground for the workforce technology platform (Case 01), feeding it in a direct loop — every model we tested became a requirement the technology had to flex around, and every process that proved critical became part of the standardization floor the platform locked in.
Standardize the questions, not the answers.
What this page doesn't show
The operating playbook — how the work was sequenced, how champions were won, and how the change was made to stick. That part isn't a storyboard. It's a conversation.