Keep Your DME Billing System. Retire the Copy-and-Paste That Feeds It.
Most suppliers who want to automate fax intake assume it means a new platform, a long migration and a retrained team. It usually doesn't. What decides whether automation pays off is how far the data travels into the system you already run, and whether a person checks it on the way in.
Why this matters right now:A lot of DME teams set next year's budget and renew software contracts in the last weeks of the year. Intake vendors are pitching AI hard, and some billing platforms now sell their own document automation add-ons. Before you sign anything, find out which level of integration you're actually buying. "Integrates with Brightree" can mean anything from a PDF in a shared folder to clean, reviewed fields on the right patient's order.
The fear that keeps fax intake manual
Ask an intake manager why referrals are still keyed by hand and you'll often hear some version of this: "We can't change billing systems right now." Fair enough. Your billing platform holds years of patient history, payer setups, rental schedules and resupply cycles. Moving off it is a project nobody wants in the middle of a busy year.
Intake automation doesn't need to touch any of that. Faxes still arrive the way they do today. Your billing system stays the system of record. The only part that changes is the stretch in between, where a coordinator opens a 20-page fax, works out what's in it, renames the file, finds the patient and types the order details in by hand. That stretch is where most of the time goes, and it's where most of the errors start.
One referral, start to finish, done by hand
Here's a hospital bed referral on a Tuesday afternoon. The fax is 22 pages: a cover sheet, a standard written order, two visit notes, a face sheet, an insurance card scan and a few pages nobody asked for. Your coordinator scrolls through it, splits out the order and the notes, and saves them as something like "Smith_J_bed.pdf".
Then she opens the billing system and searches for the patient. There are two John Smiths with similar birth dates, so she makes a new chart to be safe. She types the date of birth, the Medicare number from a grainy card image, the ordering physician's NPI, E0260 and the diagnosis code. She uploads the PDF to the new chart and moves on to the next fax.
Nothing in that story is careless. It's how most desks run. But look at what could go sideways. One wrong character in the Medicare number and eligibility bounces. The NPI she copied might be the group's and not the individual ordering physician's, which turns into a PECOS mismatch later. The extra chart means the next resupply order lands on the wrong record. And the visit notes that support medical necessity are now attached to a chart billing won't look at.
The JAMIA study above looked at clinicians copying glucose readings, which is a much simpler job than reading a faxed order. If they got 3.7% of entries wrong, it's hard to believe a busy intake desk keying a dozen fields per referral does better.
"Every field in your billing system got there through somebody's keyboard, usually on a day when the fax queue was 40 deep."
What "integration" can mean, from thin to deep
Vendors use the same word for very different things. This is the scale we use when suppliers ask us how to compare them.
| Level | What happens to the fax | What your team still does by hand | Error exposure |
|---|---|---|---|
| 1. File drop | Fax converted to PDF and saved to a folder or inbox | Everything: sorting, naming, patient lookup, keying, uploading | High |
| 2. Classified and split | Pages sorted into order, notes, face sheet, insurance and so on | Patient match, keying every field, uploading | High |
| 3. Attached to the right record | Documents matched to the existing patient and order in your billing system | Keying demographics, payer, NPI, HCPCS and ICD-10 | Moderate |
| 4. Fields written back, with review | Key fields extracted, checked by a person, then written into the billing system | Reviewing flagged fields and fixing exceptions | Low |
| 5. Two-way status | Missing items (signature, notes, face-to-face) tracked and pushed back to intake and the referral source | Working the exception queue | Low |
Levels 1 and 2 save some scrolling, but the coordinator still re-keys the referral. Level 3 is where wrong-chart problems start to go away. Level 4 is where re-keying stops. Level 5 is what gets you out of the "did we ever get the signed order back?" spreadsheet.
Pay attention to the "with review" part of level 4. Writing extracted fields straight into a billing system with nobody checking them just moves the error from a person's keyboard to a machine's guess. A good setup shows the reviewer the field and the spot on the page it came from, and it flags anything it isn't sure about.
What the re-keying is costing you
The labor cost is the easy part to see. If a coordinator spends eight to ten minutes keying and filing each referral, a desk doing 150 referrals a day is spending more than 20 staff hours a day on typing. That's before anyone chases a missing signature.
The bigger costs show up later and somewhere else. Eligibility rejections from a mistyped Medicare number. Denials for an ordering provider who doesn't match PECOS. Duplicate charts that split a patient's history, so the same-or-similar check misses an earlier purchase. And audit requests where the supporting notes exist but sit on the wrong record, so nobody can find them before the ADR deadline.
None of those land on the intake team's report. They show up in the denial queue weeks later, and the link back to a keying slip on a Tuesday afternoon is usually lost.
Where the manual hand-off breaks
The blue steps are slow, but they don't do much damage. The red ones are where the money goes. Automation that stops at step 2 leaves the expensive part alone.
How to roll it out without a migration
Start in shadow mode. Let the automation read the same faxes your team is already working, and compare its output to what your coordinators keyed. Two weeks of side-by-side results tells you more than any demo.
Then go live with attachment only. Documents get matched and filed to the right patient and order, and your team keeps keying fields. This fixes the wrong-chart problem first and builds trust with the people who'll use it every day.
Turn on field write-back last, one field group at a time. Demographics and payer first, then ordering provider, then HCPCS and diagnosis. Keep a person reviewing anything flagged as low confidence, and watch your eligibility rejection and duplicate chart rates as you go. If they don't drop, something's mapped wrong, and you'll find it in days rather than in a quarter-end denial report.
Integration checklist
You can get through most of these questions in a single call. The answers will tell you quickly whether a product works with your billing system or just sits next to it.
Automate intake without changing billing systems
DocuFindr reads every incoming fax, sorts and splits the packet, matches it to the right patient and order, and sends reviewed fields into the billing system you already run. Your team works exceptions instead of typing. Let's look at what that would mean for your desk.