The One-Page Referral Letter a Physician Acts On

You noticed something in session that needs a physician — a voice that has stayed hoarse for six weeks, a four-year-old who cannot cross midline and keeps failing the same visual tasks, a knee that is not behaving like a musculoskeletal knee. You write a careful letter: history at the top, your evaluation summarized, findings in the middle, and somewhere in the final paragraph, softly, the reason you wrote. Then nothing happens. The problem is rarely that the physician disagrees with you. It is that the letter arrived in an inbox that gets processed in seconds per document, and it was built to be read like an essay when it needed to be built to be triaged. This article rebuilds the therapy referral letter around how it is actually consumed: the ask in the first line, findings before history, one page total — with a copy-ready version for a specialist and a different one for a pediatrician, because those are two different letters.

The research on referral communication — most of it studying letters between primary care and specialists — keeps finding the same two failure modes, and both transfer directly to therapy letters. First, missing clinical information: when the letter omits the findings, the relevant history, or what has already been tried, the receiving clinician cannot triage it, and the patient waits while the information is chased or the visit starts from zero. Second, unclear purpose: reviews of referral quality consistently identify the question being asked of the specialist as essential content, and it is the element most often buried or missing entirely. Structured letters and templates, the same literature finds, measurably improve what makes it onto the page.

Therapists have an extra handicap in this exchange: the physician knows what a cardiology consult note is for, but may have only a vague model of what an SLP, OT, or PT actually measured. So a therapy referral letter carries a double load — it has to make its ask, and it has to make your data legible to a reader who does not use your instruments. That is an argument for translation, not for length. Report the finding with its functional meaning in the same sentence: not a raw test score, but the score plus what the child can and cannot do because of it. Every finding earns its place on the page by being something the physician can act on or weigh; everything else is drag.

The structural fix is a single inversion. Therapists write chronologically, the way an evaluation report is organized: referral source, history, testing, findings, and finally impressions. A physician processing an inbox reads in the opposite order of importance: What do you want from me? How urgent is it? What is the evidence? A letter that answers those three questions in the first four lines gets acted on. A letter that makes the reader excavate for them gets set aside for later, and later has a way of not coming.

If writing to a physician feels presumptuous, the professional associations disagree. APTA’s Standards of Practice state it flatly: when the diagnostic process reveals a need for care outside the scope of physical therapy, the therapist informs the patient and refers them to an appropriate practitioner — a duty that grows heavier as direct access makes PTs a true first contact. AOTA’s Standards of Practice direct occupational therapists to recommend consultations or refer when the client’s needs are best served by another professional’s expertise. ASHA’s Code of Ethics requires clinicians to use every resource, including referral and interprofessional collaboration, to ensure quality service, and its Scope of Practice names collaboration as a core domain. The referral letter is the instrument that converts this obligation into an appointment. An undocumented hallway suggestion to “maybe get that looked at” protects no one — not the patient, and not your chart.

1. 01

   The ask, in the first line

   Name the patient, name what you are asking for, and say why in one clause: “I am referring [patient] for evaluation of [finding].” If there is urgency, it goes here too — not as alarm, but as a timeframe. The reader should be able to triage the letter without reading past line three.

2. 02

   The findings that justify it

   Three to five observations, most significant first, each translated into function. Dates, measures, and what changed over time. This is the evidence section, and it is the only part of the letter where your clinical data belongs.

3. 03

   Only the history that changes the picture

   Relevant diagnoses, relevant medications if known, what has already been tried in therapy and how the patient responded. Leave out the full birth history and the complete evaluation — offer to send the report instead of attaching your whole chart.

4. 04

   The close: logistics and the loop

   What you will keep doing in the meantime, how to reach you directly, and an explicit invitation to send findings back. A referral letter that does not ask for a reply has volunteered to be a dead end.

Everything below one page is a cut, not a loss. The evaluation report still exists; the letter’s job is to get the right appointment made, and the report’s job is to be there when the specialist wants depth. Conflating the two documents is the root mistake — a letter that tries to be the report gets treated like a report: filed, unread.

Write to a specialist when the path is direct — the family can book an ENT, orthopedist, audiologist, or developmental evaluation without a gatekeeper, or the specialist’s office has asked for documentation to support scheduling. The specialist letter is the more technical of the two: this reader shares vocabulary with you at the edges of your scopes, wants the ask narrow, and wants the findings dense. You are not asking them to manage the patient; you are asking them to answer a question.

Write to the pediatrician — or any primary care physician — when the patient needs a medical home’s judgment, not a named specialist’s procedure. Sometimes that is because the insurance plan routes specialist visits through the PCP; often it is because the pediatrician has known this child for years and holds context you do not. The letter changes accordingly. You are not directing traffic; you are contributing your data to a decision the pediatrician will make, and the ask softens from “please evaluate for X” to “these findings suggest X is worth evaluating — I defer to your judgment on the path.” The deference is not politeness. Whether a payer requires a PCP referral for the specialist varies by plan, and the pediatrician may know things — a recent well-child visit, a family situation, a prior workup — that resolve your finding without a new appointment.

The single highest-leverage edit in any referral letter is converting a topic into a question. “Concerns about swallowing” is a topic; the reader cannot finish it. “Is there an anatomical cause for the coughing during thin liquids that should change my dysphagia plan?” is a question; the reader can answer it, and answerable questions get answered. The test: could the physician, in one sentence of reply, resolve your letter? If not, the ask is not yet explicit.

The letter is PHI in motion, so it travels by a channel your practice has vetted — fax, a secure direct-messaging address, or a portal — not by whatever email is handy. Then it goes in your own chart: the referral, the date, the recipient, and the family conversation about it. That entry is doing real work. If the finding you flagged turns out to matter, the chart shows you acted on it the day you saw it; and when the specialist’s report comes back, the entry is the hook it hangs on. The letter also earns its keep at the other end of the episode — the physician who received a clear referral letter mid-treatment, and a progress summary when you promised one, is the physician whose office remembers your practice when the next family asks where to go for therapy.

Silence is the most common outcome to plan for. Physician offices lose faxes, inboxes bury letters, and families sit on referrals — none of it is a verdict on you. Give the loop a default: a calendar entry two to three weeks out to check with the family whether the appointment happened, and one polite resend or call to the office if it never arrived. For a finding you judged urgent, the follow-up is not a courtesy but part of the referral itself, and your chart should show it. What the letter cannot do is force the outcome — the family may decline, and the physician may disagree with your read. Both are allowed. Your obligations are to communicate the finding clearly, document that you did, and adjust therapy to whatever answer comes back, including no answer.

“

“A referral letter is a handoff, and a handoff is judged the way all handoffs are: not by how much you carried, but by whether the other person caught it.”