A Practitioner Without a Working Device Is Care You Can't Deliver

Multi-site healthcare has a structural problem that single-site businesses never see: the fleet is everywhere and the IT team isn't.

A regional operator with twenty or forty locations typically runs endpoint support with a handful of people — sometimes two. Every clinic has front-desk machines, exam-room devices, practitioner laptops, and a nurse station or two. Every one of those endpoints is load-bearing: the schedule, the chart, the orders, the billing all flow through them. And when one dies, the person best positioned to fix it is, on an average day, a car ride away — behind three other tickets.

So the real recovery model at most multi-site operators, whatever the documented process says, is this: the device fails, the site calls IT, IT triages by phone, someone drives or ships something, and the affected practitioner or front-desk person works around the gap for hours or days. Everyone involved is competent. The model is what fails.

The schedule is the meter

Healthcare downtime has a property most industries don't: the cost is metered in appointment slots, and the meter is public.

Run the illustrative arithmetic on a single incident. A provider working twenty-minute slots sees roughly three patients an hour. A device failure at 9am that isn't resolved until early afternoon is twelve to fifteen visits that either compress into an already-tight schedule, rebook into a calendar that's full for weeks, or simply don't happen. Each of those is deferred or lost reimbursement — and a patient whose care slipped because of a laptop.

Front-desk failures are worse per minute, because check-in is serialized: one dead machine backs up every patient in the lobby, for every provider in the building.

None of this appears in an IT metric. Ticket systems will record the incident as resolved-within-SLA. The schedule records it differently.

Why fix-in-place fails across distributed sites

The instinct is to get better at repair: faster remote tools, better spare parts, more phone triage. Across a distributed footprint, this optimizes the wrong step. Diagnosis-then-repair puts the slowest, least predictable work — figuring out what's wrong — inside the window where a clinician is waiting. It also assumes technician proximity that multi-site operators, by definition, don't have.

The model that fits the physics is swap, then diagnose:

• A known-good replacement device — imaged for that role, ready to sign in — reaches the site fast, because it was staged for exactly this.
• The swap is minutes of work a site coordinator can do with guided support. The practitioner is back in the schedule.
• The failed device ships to a bench, where diagnosis happens without a clinician waiting on it. It's repaired or retired, wiped with certification either way, and returns to the pool.

Time-to-working-device stops being a function of technician drive time and starts being a function of logistics — which, unlike drive time, can be engineered.

The part healthcare can't skip: what happens to the old device

In most industries, a dead laptop in a drawer is waste. In healthcare, it's exposure. Any device that has touched clinical workflows has to be presumed to carry PHI, which means the back half of the swap — recovering the failed unit, wiping it, and proving both — is a compliance obligation, not housekeeping.

That's why device recovery in healthcare has to be a custody operation, not a shipping errand:

• Serialized tracking from the moment the replacement ships and the failed unit is boxed — so the asset record always matches reality, site by site.
• Documented chain of custody for the return leg. "It's somewhere between the clinic and us" is not an answer you want to give a privacy officer.
• Certified data destruction on every retired or redeployed unit, with the certificate filed against the serial number — so when the audit or the security questionnaire asks, the evidence already exists.

Do this as a byproduct of normal operations and audits become file-retrieval. Do it ad hoc and every departed laptop is an open question with PHI implications.

What good looks like

For a multi-site healthcare operator, an engineered device operation means: every endpoint serialized and tracked; role-based known-good configurations maintained for practitioners, front desk, and nurse stations; replacements staged so time-to-working-device is measured in minutes-to-hours, not days; failed and retired units recovered on a documented custody chain and wiped with certification; and your two-person IT team pulled out of the car and back onto the work only they can do.

The clinics keep seeing patients. The schedule stops noticing IT.

See how we run healthcare fleets