TL;DR: To automate incoming faxes into Greenway Intergy patient charts, route your fax lines into an AI agent that classifies each document, extracts the patient's name and date of birth, matches the record, and files the page to the chart with the right category and task. Start with one fax line, set a confidence threshold so uncertain matches go to a person, and track turnaround time and misfile rate for a few weeks before you expand.
What does it mean to automate faxes into Intergy charts?
Automating faxes into patient charts means software does the reading, matching, and filing that your staff do by hand today. A fax arrives, the system works out what it is and whose chart it belongs to, and the document shows up in that chart already categorized, with a task waiting for the right person.
It doesn't mean removing people from the process. It means people stop touching the routine majority of faxes and spend their time on the ones that actually need judgment. The size of that majority varies by practice, so measure it during your pilot rather than assuming it.
If you run Intergy, you likely already have some of the plumbing. Greenway Document Manager, built on the ProviderFlow platform and supporting Intergy and Prime Suite, includes scanning and eFax. Those tools get documents into a digital queue. Automation picks up from there and handles the indexing.
Step 1: Audit your current fax volume and document mix
Before you automate anything, find out what you're automating. Pull two weeks of fax data and answer these questions:
- How many faxes arrive per day, and how many pages is that?
- How many separate fax numbers or inboxes feed the practice?
- What are your top five document types by volume?
- How long does a fax sit before someone files it?
- Who touches each document, and how many minutes does it take?
The last two matter most. Turnaround time tells you what delay is costing you, and minutes per fax tells you what the labor is worth. If a referral waits two days in the queue, you're losing scheduling slots. If it takes three minutes of staff time to index a lab result, multiply that by your weekly volume and you have your baseline.
Specialty practices usually find that referrals and outside records dominate. Primary care groups tend to see more labs, imaging reports, and refill requests. Knowing your mix tells you where to point the pilot.
Step 2: Centralize your inbound fax lines
Automation works best when faxes arrive in one place. If you have four fax numbers ringing at four different desks, consolidate them into a cloud fax service or a shared digital inbox first.
This step is dull and it pays off. A single entry point gives the triage software one source to monitor, gives you one place to measure volume, and removes the "which desk got that referral?" problem. It also doesn't require you to change anything for the people sending you faxes. Referring offices keep using the number they've always used.
Keep the HIPAA basics in mind here. Any vendor that handles faxes containing protected health information is a business associate, and HHS publishes sample business associate agreement provisions that your agreement should align with. Get the agreement signed before a single real fax flows through.
Step 3: Define classification and routing rules
This is where you decide what the automation does with each type of document. Write it down as a simple table with three columns: document type, where it files in the chart, and who gets the task.
A starting version for a specialty practice might look like this:
- New referral: file under referrals, create a task for the referral coordinator.
- Lab or imaging result: file under results, route to the ordering provider.
- Prior authorization response: file under insurance, route to the authorization team.
- Outside records: file under outside records, route to the care team for review.
- Refill request: route to the refill queue.
- Unclassified or non-clinical: send to a human review queue.
Be specific about urgency. Decide in advance which keywords or document types should jump the queue, such as a critical lab value or a stat referral, and where they go. Then write a rule for what happens when the system isn't sure, because it won't always be.
The routing table is the part most practices underestimate. The AI can only send work where you tell it to, so a vague table produces vague results.
Step 4: Set confidence thresholds and a human review queue
Every classification and every patient match comes with a confidence level. Your job is to decide where the line sits.
A sensible approach is to set it conservatively at first. Above the threshold, the document files automatically. Below it, the document lands in a review queue where a staff member confirms or corrects it. As you see how the system performs on your faxes, you can lower the threshold for document types where it's consistently right.
Pay particular attention to patient matching. Classifying a document as a lab result is low-risk if it's wrong, because someone will notice. Filing it to the wrong patient is a privacy problem. Most practices want a stricter threshold for patient matching than for document type, and many require a match on at least two identifiers, such as name and date of birth.
The review queue isn't a failure state. It's the part of the design that makes the rest safe. Assign an owner, set a service level for clearing it (same day works for most practices), and review what lands there each week. The patterns tell you what to fix.
Step 5: Pilot, measure, and expand
Start narrow. Pick one fax line and one or two document types, referrals being the usual choice, and run the automation for three to four weeks while your team keeps its normal process as a backup for the first few days.
Track four numbers against your baseline from Step 1:
- Turnaround time from fax received to filed in the chart.
- Auto-file rate, meaning the share of faxes that needed no human touch.
- Misfile rate, from spot checks on a sample each week.
- Review queue volume and how long items wait.
If turnaround drops, the misfile rate stays at or below your manual rate, and the queue stays manageable, expand to the next document type or fax line. If not, adjust the rules or thresholds before you add volume.
On the people side, tell your front desk what's changing and why. Staff turnover is a live concern: MGMA's May 2026 poll found 28% of practices reporting higher turnover than the year before, with front-desk and administrative roles among those losing the most people. Framing automation as taking the filing work off their plate, not replacing them, matters for adoption.
Where does fax automation go wrong?
Most failures come from a handful of predictable problems. Know them before you start.
- Poor scan quality. Faint, skewed, or low-resolution pages reduce accuracy. Some of this you can't control, since you don't control how other offices send faxes.
- Patient-match ambiguity. Two patients with the same name, a married name that changed, or a fax with no date of birth. These belong in the review queue.
- Multi-document faxes. One transmission containing a referral, an insurance card, and a lab result needs to be split before it can be filed correctly.
- Handwriting. Handwritten referrals and notes remain the hardest category. Expect more of them in review.
- Skipped rule-writing. Teams that buy the tool but never define their routing rules end up with filed documents that nobody sees.
None of these are reasons not to automate. They're reasons to run a pilot, keep a review queue, and measure.
Platforms like Honey Health offer AI agents for this workflow, including fax triage that classifies incoming documents, matches them to patients, and files and routes them while your EHR stays the system of record. Whichever tool you pick, insist on a trial with your own faxes.
Frequently Asked Questions
Can you automatically file faxes into Greenway Intergy?
Yes, with the right tooling. Greenway's own document tools handle scanning and eFax, and AI fax triage software can add classification, patient matching, and routing on top. How documents land in the chart depends on your configuration and the vendor's integration method, so ask for a demonstration in a test environment.
How accurate is automated fax indexing?
Accuracy depends on fax quality, document type, and the cleanliness of your patient data, so any single number from a vendor should be tested against your own sample. Run 200 to 300 of your recent faxes through the system and measure classification accuracy, patient-match accuracy, and the share sent to human review.
Do I need to change my fax number?
Usually not. Most setups forward your existing fax number or inbox into the automation layer, so referring offices and labs keep sending faxes the way they always have. Confirm the setup with your vendor and your fax provider.
What happens to faxes the system can't read?
They go to a human review queue. Well-designed systems flag low-confidence documents instead of guessing, so a staff member can confirm the patient and document type. Assign an owner to the queue and set a same-day target for clearing it.
How much staff time does fax automation save?
It depends on your volume and how long indexing takes today. Multiply weekly faxes by minutes per fax to get your current labor cost, then subtract the time spent on review-queue items. Measure it during your pilot rather than relying on an industry average.
