TL;DR: You set up automated inbound fax triage with CureMD by routing your existing fax line through an AI layer that reads, classifies, and files documents into CureMD — no new fax number and no EHR rebuild required. The practical path is four steps: connect the fax feed, define your document categories and routing rules, map each document type to the right CureMD chart location, then set a human-review threshold and monitor accuracy as the system tunes to your document mix.
What setting up fax triage with CureMD involves
Setting up fax triage doesn't mean ripping anything out. Your CureMD instance stays as-is, your fax number stays as-is, and referring offices keep sending to the same place. What you're adding is a layer in between: software that catches every inbound fax, figures out what it is and who it's for, and files it into CureMD so staff stop doing that by hand.
That distinction matters because fax isn't going away. It's still around 75% of all medical communication, and the MGMA has documented how stubbornly it persists in medical groups. So the goal of setup isn't to eliminate faxes — it's to make the inbound pile handle itself.
The whole process usually takes days, not months, because the automation runs on the fax stream rather than deep inside the EHR. Here's the sequence.
Step 1 — Connect your fax feed to the triage layer
The first move is getting every inbound fax into one digital channel the triage system can read. You have two clean options: port your existing fax number to the triage platform, or forward it there from your current fax or e-fax service. Either way, nothing changes for the offices sending to you.
This is the step where the platform choice shows up. A tool like Honey Health's Fax Triage agent sits directly on the inbound fax line, so once the feed points at it, every incoming document flows through classification and filing automatically. Because it works off the fax stream, you don't need CureMD to expose a deep API or build a custom integration — the automation reaches CureMD to file finished documents, but it doesn't depend on the EHR to do the reading.
Confirm one thing before you move on: that every fax source is captured. If you have multiple fax numbers across locations, route them all in now so nothing sneaks past the automation later.
Step 2 — Define document categories and routing rules
Next, tell the system what kinds of documents you receive and where each should go. Most practices land on a category set like this:
- Referrals — inbound referral packets that need to reach the referrals or scheduling team.
- Lab and imaging results — clinical results headed for the ordering provider's queue.
- Records requests — release-of-information and outside-records documents.
- Prior authorization and insurance correspondence — payer documents that belong with billing or the PA team.
- Everything else — a catch-all so nothing silently disappears.
For each category, you set a routing rule: which internal queue it goes to and where it files in CureMD. Keep the first version simple. You can always add finer categories once you see your real document mix — over-engineering the taxonomy on day one just slows the launch.
Step 3 — Map document types to CureMD and set a review threshold
With categories defined, map each one to its destination in CureMD — the chart location and the staff queue that owns it. This is what turns "the system knows this is a lab result" into "the lab result is filed to the right patient's chart and the ordering provider sees it."
Then set a confidence threshold. The AI scores how sure it is about both the document type and the patient match. Anything above your threshold files automatically; anything below routes to a human review queue instead of getting filed on a guess. Start conservative — a higher threshold sends more documents to review early on, which feels slower but builds trust while you confirm the system is matching correctly. As accuracy proves out, you lower the threshold and more files clear on their own.
Step 4 — Monitor accuracy over the first weeks
Fax triage isn't a set-and-forget switch on day one; it's a system that tunes to your practice. For the first couple of weeks, watch three things: how often the system classifies correctly, how often it patient-matches correctly, and what's landing in the review queue and why.
The review queue is your feedback loop. If a specific referring office's cover sheet keeps confusing the classifier, or a particular document type keeps missing its patient match, that's a tuning signal — adjust the rules and the accuracy climbs. Within a few weeks most practices settle into a rhythm where staff touch only genuine exceptions, and the routine majority of faxes file themselves. Speed matters here too, because 88% of practitioners say fax-related delays affect patient care — the faster a referral clears triage, the faster the patient gets booked.
Common setup pitfalls to plan for
A few predictable issues trip up first-time setups. Plan for them and the launch goes smoothly:
- Multi-page faxes with several documents. One fax transmission often contains a referral, a note, and a lab result stapled together. The system needs to split these into separate documents; confirm document-splitting is on and working early.
- Unmatched patients. New patients or misspelled names won't match an existing CureMD chart. Decide up front where these route — usually a dedicated queue where staff create the chart, not an auto-file.
- Cover-sheet noise. Fax cover sheets and marketing junk can confuse a classifier. A good setup learns to ignore them, but expect a short tuning period.
- Orphaned fax numbers. If one location's fax line never gets routed into the system, its faxes keep piling up manually. Audit every number during Step 1.
Handle these during setup rather than after go-live, and the transition is nearly invisible to your team.
Frequently Asked Questions
Do I need a CureMD API or custom integration to set up fax triage?
No. The automation runs on your inbound fax stream, so it classifies and files documents into CureMD without requiring a deep native integration. That's why setup takes days rather than the months an EHR-level integration project would.
Will we lose any faxes during the switchover?
Not if you route every fax source into the system during setup. The safe approach is to forward your existing numbers so faxes continue arriving normally while the automation begins handling them — there's no gap where faxes go uncaptured.
How long until the system is accurate?
Most practices see strong accuracy within the first few weeks as the classifier tunes to their document mix. Starting with a conservative review threshold means low-confidence documents go to staff early, so nothing is misfiled while the system learns.
Can we still file some faxes manually?
Yes. Anything the system routes to the human-review queue is handled by staff exactly as before. Over time that queue shrinks to genuine exceptions, but you always keep a manual path for the documents that need judgment.
Does setup disrupt our current CureMD workflows?
Minimally. Documents still end up in CureMD where your team expects them — the change is that they arrive already classified and filed. Staff shift from sorting every fax to reviewing only the exceptions.
