The 6 Places DME Companies Leak Money (And the Software That Plugs Each One)

I spent a few years doing operational due diligence on DME companies — mostly for buyers, occasionally for owners who wanted to know what a buyer would find. It is a weirdly intimate job. You see the real numbers, not the ones on the website.

And after enough of them, you start noticing that the leaks are always in the same six places. Different companies, different sizes, different states, same six holes. It stopped being interesting and started being almost funny.

The other pattern: almost nobody knows which of their six holes is the biggest. Owners have a strong intuition, and the intuition is wrong about half the time. The loudest problem is rarely the most expensive one.

So here’s the list, roughly ordered by how much money I typically found in each, with what actually fixes it.

1. Documentation collected too late (usually the biggest)

This is the champion. Every time.

The pattern: the order gets taken, equipment goes out, and somewhere in the following weeks someone discovers a required document was never collected. Now you’re chasing a physician’s office for a signature on something that happened three weeks ago, and you are not their priority. Some of those you win. A meaningful percentage you write off.

Nobody books this as a loss. It shows up as “aged AR” and then as an adjustment, and it’s invisible as a category. When I’d add it up as its own line, owners would go quiet.

What fixes it: a system that knows what documentation this product needs for this payer and won’t let the order proceed without it. Not a checklist someone might use. An enforced gate at intake.

This is the single strongest argument for a unified platform, and it’s why NikoHealth tops my list overall — intake, documentation, delivery, and billing in one system means the front end knows what the back end will need, and enforces it at the moment it’s cheap to collect. Their revenue cycle management software approach is built around that prevention idea rather than around better tools for cleaning up afterward, which is the right instinct.

2. Inventory you own and cannot find

Second biggest, and the most emotionally upsetting for owners, because it’s physical. It’s right there. Somewhere.

Three flavors, all common:

Equipment at patients’ homes that stopped billing. The rental ended, nobody picked it up, and it’s now a $900 asset living in a stranger’s garage generating nothing. I have found six figures of this at mid-size companies. Not once. Repeatedly.

Stock nobody can see, so you buy it twice. Classic multi-location failure. Branch A orders what Branch B has on a shelf.

Equipment stuck in repair limbo. It went to be fixed, the work order died, and it’s been sitting in a corner for eight months.

What fixes it: serialized tracking that follows the asset through every state — warehouse, truck, patient, repair — with real-time visibility across sites. This is precisely what dme inventory software should mean, and the crucial part is that it lives inside the order system rather than beside it. A standalone warehouse tool doesn’t know a rental ended. The order system does.

Cheap thing you can do Monday with no software: pull every rental that stopped billing more than 90 days ago and hasn’t been picked up. That report has paid for itself in every single engagement I’ve run it in.

3. Proof of delivery that doesn’t come back

Smaller per incident, brutal in aggregate, and the most fixable thing on this list.

The equipment was delivered. The patient got it. But the signature is on a sheet in a van, or a photo on a tech’s phone, or genuinely just gone. Billing can’t submit, or submits and loses the audit.

I once traced a company’s entire denial spike to one tech who’d been out for six weeks and whose paperwork nobody had processed. Six weeks of deliveries, unbilled, discovered in month three.

What fixes it: mobile capture that attaches the POD to the order automatically at the moment of delivery. Not “the tech uploads it later.” At the door, on the device, attached. If your proof of delivery has a manual step between the driveway and the claim, you will lose some of it forever.

4. Authorizations that expire in the gap

The auth was obtained. Then the order sat — backorder, scheduling, patient unavailable — and by delivery day it had lapsed. Nobody re-checked, because nobody’s job was to re-check.

This one is sneaky because the paperwork looks complete. You did get the auth. It’s just not valid anymore.

What fixes it: automated tracking with alerts tied to the actual order timeline, not a spreadsheet with dates in it. A spreadsheet doesn’t know the delivery slipped two weeks.

5. Denials that are never worked

Not denials you lose. Denials nobody touches.

Every operation I’ve reviewed had a bucket of claims that were denied, appealable, and simply aged out because the worklist was long and these were at the bottom. Usually the mid-value ones — big enough to matter, not big enough to be anybody’s emergency.

What fixes it: workqueues prioritized by dollars and by appeal deadline, not by date received. And, honestly, fixing hole #1, because a team drowning in preventable denials will never get to the bottom of the list.

6. Routes built by vibes

The smallest of the six, but it’s pure margin and it’s constant.

Most DME delivery routes are built by a dispatcher’s mental map. Those people are often genuinely excellent — and they’re also a single point of failure who can’t account for what’s on which truck, who’s closest right now, and which stops can be combined.

What fixes it: route optimization that’s connected to inventory and order readiness. Standalone optimizers produce beautiful routes to the wrong stops, because they don’t know an order isn’t ready.

How to find your biggest hole in one afternoon

You don’t need a consultant. Here’s the version I’d hand you for free:

Pull 50 denials from last quarter, at random, and trace each to root cause. Not the denial code — the actual originating mistake, and which department made it. An afternoon of work. It will surprise you.

Pull rentals that stopped billing 90+ days ago with no pickup logged. Multiply by replacement cost. That’s a number you’re currently not seeing anywhere.

Count days from delivery to claim submission. If it’s more than two or three, your PODs are sitting somewhere. Go find where.

Ask your dispatcher what happens if they’re sick. If the honest answer is “things get bad,” you have a person-shaped dependency where a system should be.

The uncomfortable conclusion

Five of these six leaks are upstream of the department that gets blamed for them. Billing gets blamed for #1, #3, #4 and #5. Billing caused approximately none of them.

Which is why I’ve become boring on this subject: the answer is almost never a better tool for the department that’s suffering. It’s a system where the mistake can’t be made two steps earlier.

Go run the 50-denial trace. Seriously. It’s one afternoon and it will tell you more about your business than any dashboard you own.