Device Enrollment Logistics for Pulmonology Practices

A pulmonologist places the order and considers the matter closed. Six weeks later the patient calls asking when the machine is coming, and nobody at the practice can say, because device enrollment logistics is the one part of this workflow that belongs to no department.

The order leaves the practice and enters a space with three other parties in it: a supplier, a plan, and a patient. Each can stall the process, none reports back on a schedule, and the practice finds out something went wrong only when the patient complains or when the follow-up visit arrives and the device has never been used.

By then the delay is not a delay. The patient has spent two months without therapy they were prescribed, the practice looks disorganized, and a chunk of the visit that was supposed to review progress gets spent reconstructing what happened to the order.

The coverage rules are the reason this takes so long

Practices tend to describe the wait as supplier slowness. Most of it is documentation, and the requirements are published.

Medicare pays a supplier to rent a CPAP machine for 13 months of continuous use before the patient owns it, and after the Part B deductible the patient pays 20% of the approved amount for the rental and the supplies. That structure means the supplier carries risk across more than a year, so they will not ship until the file is complete, and an incomplete file sits rather than bouncing back.

Medicare’s durable medical equipment coverage rules also require the equipment to come from a supplier enrolled in the program, which is why a patient who found their own vendor sometimes cannot use it. That is a fact worth catching at order time rather than three weeks in.

None of this is clinical. It is a document set, a supplier relationship, and a coverage check, and every one of those is administrative work the front office can track, chase, and close.

The order queue is the tracking system nobody works

The information needed to run this well already exists inside athenaOne. It just has no owner.

Reading open orders through GET /chart/encounter/{encounterid}/orders gives the practice a live list of every device order placed and not yet resolved, with the date it was written. That list is the entire program. Sorted by age, it shows which patients have been waiting longest, and the ones at the top are almost always the ones nobody has called.

Creating a tracking case with POST /patients/{patientid}/documents/patientcase against the right department bucket turns each of those into a piece of work with a status rather than an open-ended hope. Cases carry the attempt history, so the fourth call knows what the first three found, which is the difference between a chase and a loop.

The practical version is a standing sweep. Orders older than a set number of days with no confirmed delivery get contacted: the supplier for status, the patient to confirm they were reached and that the address and phone are right. Most stalls resolve on the first call, because the common failure is that the supplier called once, got voicemail, and moved on.

The complication: the patient does not know they are the bottleneck

This is the failure mode that surprises practices, and it accounts for a large share of the stalled orders.

The supplier calls from a number the patient does not recognize, leaves a message about equipment, and the patient assumes it is a sales call. Or the supplier needs the patient to confirm delivery details and the patient is waiting for the practice to tell them what happens next, because the practice is who they trust. Both sides believe the other has the ball, and the order ages quietly.

An outbound call that names the practice breaks that stalemate immediately. Reading the patient record through GET /patients gives the right contact details and language preference, and the call does three administrative things: confirms the patient has heard from the supplier, confirms the address and phone the supplier is using are current, and tells the patient plainly what the next step is and who will call.

The handoff is firm and it is worth stating clearly. If the patient asks whether they still need the device, whether they can stop using it, or anything about their symptoms or how the therapy is going, the automation does not answer. It creates a case routed to clinical staff with the question captured verbatim. The AI closes logistics gaps. It does not discuss the therapy.

Enrollment is not adherence, and conflating them creates scope problems

Device programs have a second phase that looks similar and is governed by completely different rules.

Enrollment logistics ends when the device is delivered, set up, and the patient knows how to reach the supplier. Everything after that, whether the patient is using it, how many hours a night, whether the numbers coming back are acceptable, is clinical territory. Compliance thresholds exist in coverage policy and they are read and acted on by clinicians, not by a front-office system.

The useful boundary is concrete. Administrative work in the second phase looks like booking the follow-up visit the coverage rules imply, calling patients whose supplier reports no contact, chasing a resupply order that was written and never filled, and confirming the patient still has the supplier’s number. None of that requires interpreting a single reading.

Practices that blur this end up with an automation asking patients how their therapy is going, which is both out of scope and unhelpful, because the answer has to go to a clinician anyway. Keeping the line clean makes the tool more useful, not less: the front office reliably closes the logistics, and the clinical team gets clean handoffs instead of a transcript to sort through.

What to track, starting with the number nobody has

Almost no pulmonology practice can state the median time from order written to device delivered. That single number is the program.

Measure it, then break it into segments, because the fixes are different. Time from order to supplier acknowledgment is a routing problem. Time from acknowledgment to documentation complete is a paperwork problem inside your own building. Time from complete file to delivery belongs to the supplier and the plan, and it is the piece you can only manage by escalating.

Then count orders older than thirty days with no confirmed delivery, as a standing figure reviewed weekly. This is the backlog, and it is usually larger than anyone expects, because there has never been a report that shows it.

The last one is the honest check. Follow-up visits where the device had not arrived or had never been set up. Those visits consume a clinical slot and produce nothing, and they are entirely preventable by a phone call made three weeks earlier. Counting them is what turns device enrollment logistics from a vague annoyance into a case someone will fund.

Key Takeaways

Work the open order queue by age. Orders sorted oldest first surface the stalled ones, and the top of that list is usually nobody’s assignment.
Create a tracked case per device order so the fourth contact knows what the first three found, instead of restarting the chase.
Call the patient in the practice’s name. A large share of stalls are a supplier voicemail the patient mistook for a sales call.
Confirm address, phone, and language preference against the record at the same time. Bad contact details are a common and invisible cause of delay.
Check supplier eligibility at order time. A patient who sourced their own vendor may not be able to use it, and finding out early saves weeks.
Stop at the clinical line. Questions about symptoms, therapy, or whether to keep using the device get captured verbatim and routed to clinical staff.
Measure median order-to-delivery time and split it into routing, documentation, and supplier segments, because each has a different fix.
Count follow-up visits where the device never arrived. That is the wasted clinical capacity that makes the business case.

Device enrollment is a coordination problem wearing a clinical costume. The prescribing decision is finished before any of this starts, and everything that follows is documentation, supplier contact, coverage checking, and telling the patient what happens next. It stays broken because it requires someone to work an aging list every week in a practice where the phones are already full. An AI team working open orders inside athenaOne can run that sweep on schedule, close the logistics gaps on the first call, and route anything clinical to the people who should be answering it.

Sources