Hours from referral received to first call
The core speed metric, and the step most likely to decide whether the patient books with you.
Physical therapy clinics · Document intake and data extraction
Get referral faxes into the schedule the same day they arrive
Physician referrals still arrive by fax at most PT clinics, often as a cover sheet, a prescription and a few pages of chart notes from an orthopedic office. Someone has to read each one, find the patient's phone number, diagnosis codes and referring provider, create the patient in the EMR and call to schedule.
How it often works today
- Faxes land in an eFax inbox or physical tray, and nobody owns them until the desk has a gap between check-ins, often late in the day.
- Staff retype name, date of birth, phone, insurance, ICD-10 codes and referring physician into the EMR, with occasional transposed digits that cause claim rejections later.
- Referrals missing a signature, diagnosis or order details go back and forth with the physician's office by fax, adding days before the first visit.
- Owners cannot see how many referrals arrived from each physician or how many became evaluations, so referral relationships are managed on instinct.
When that work waits until the desk has a quiet moment, referrals sit for a day or two and patients go elsewhere. We would set up extraction that reads incoming referrals and intake forms, fills a draft record, and leaves staff to confirm and call.
The improved workflow
Route all inbound faxes to one digital queue
Faxes arrive through a HIPAA-covered eFax service into a single queue, rather than a printer tray or a personal inbox, so nothing is lost and arrival time is recorded.
Extract key fields into a draft
A document extraction tool reads each referral and pulls patient demographics, insurance, diagnosis codes, referring provider and NPI, and order details such as frequency or protocol.
Flag missing or unclear items
Referrals without a signature, diagnosis or legible order are marked for a request to the physician's office, with a prefilled fax or portal message for staff to send.
Staff confirm and create the patient
Front desk compares the draft against the image, corrects anything wrong, creates the patient in the EMR and calls to schedule, all from one screen.
Track referral-to-evaluation time
Each referral is timestamped from arrival to scheduled evaluation, and reported by referring practice so owners see which relationships are growing.
How we deliver this: AI automation for small business or fractional CTO leadership, scoped in writing and run over 90 days. See pricing or the sample scope.
What we would measure
Referral-to-evaluation conversion
Shows how many referrals become scheduled patients, by referring practice.
Data entry corrections per referral
Tracks extraction accuracy and whether staff can trust the drafts.
Registration-related claim rejections
Wrong member IDs or dates of birth show up later as rejections, so this should fall.
Where people stay in control
- Staff confirm every extracted field against the source document before a patient record is created; nothing posts to the EMR unreviewed.
- The fax and extraction services must be covered by business associate agreements, and documents stay in approved storage rather than personal email.
- Clinical interpretation of the referral, such as protocol choice or contraindications, stays with the treating therapist.
Example
Illustrative: a clinic taking surgeon referrals
For example, a clinic receives 60 referral faxes a week, mostly post-operative knee and shoulder patients from two orthopedic groups. Each takes the desk around eight minutes to key in, about eight hours a week. With extraction prefilling drafts, review drops to two or three minutes, and referrals are called the same morning instead of the next afternoon. The illustrative gain is roughly five staff hours a week plus faster first visits for surgical patients on tight protocols.
Tools typically involved
What the first 30 days look like
In the first 30 days we would collect a sample of 50 real referrals, time the current entry process, map where faxes arrive and who handles them, test extraction accuracy on that sample, and agree acceptance criteria such as same-day first contact and correction rates.
Common questions
Can AI read handwritten referrals?
Modern extraction tools handle printed text and many typed EMR printouts well, and they can read some handwriting, but accuracy drops on poor-quality faxes and scrawled prescriptions. That is why staff review every draft. During the pilot we would measure accuracy on your own referrals and set the confidence level below which a field is always flagged.
Does our EMR already do this?
Some EMRs and fax services include document management or intake features, and some offer referral portals for physicians. We would check what your system and fax provider already support before adding anything. If a built-in option covers most of the need, that is usually the better answer than a separate tool.
What about our new-patient intake paperwork?
Digital intake forms through Jane, Prompt, WebPT or a tool like Phreesia avoid most retyping in the first place. Where patients still bring paper or PDFs, the same extraction and review approach applies. Outcome questionnaires such as the LEFS or Oswestry are often best moved to digital forms so scores land in the chart automatically.
The business case
Does it pay for itself?
Illustrative defaults for a physical therapy clinic. Replace them with your numbers.
What this workflow costs today
a month in staff time, before any lost sales
Monthly fee + assumed tools +
Break-even: new clients a month (at first-year value)
Or: staff hours recovered and redeployed
Break-even additional revenue
Value of +1 point of win rate
First 90 days, service fee
Worth a closer lookTest the assumptions
Illustration only, not a forecast or guarantee. Recovered staff time becomes a cash saving only if spending actually declines or the time goes into paid work. Tool costs are an assumption; we replace every number here with yours before recommending a scope.
Document intake and data extraction in other industries
Accounting and CPA firmsArchitecture firmsConstruction and general contractorsDental practicesElectrical contractorsEngineering firmsInsurance agenciesLaw firms
Document intake and data extraction: full guide
Related guides
Acceptance criteria for automation projectsSelf-hosting n8n for small businessHow to automate client onboarding, step by stepDocument processing automation for service businesses
Terms used here
Next step
Bring one expensive workflow. Leave with a clear next step.
In 30 minutes we look at one recurring workflow, what it costs you today and whether there is a credible case to improve it. If there is not, we will say so.
Book a 30-minute workflow review
Not ready to talk? Take the 3-minute check
30 minutes with Phil or Martin. No technical brief, no pitch deck. Phil or Martin replies personally, usually within one business day.