ClawdelleBack to blog
Occupational Therapy

AI Referral Letters for Occupational Therapists: Draft Faster Without Losing Clinical Control

How paediatric and adult OT practices can turn assessment notes into clear referral letters without automating clinical judgement.

2026-06-18 · 7 min read

Occupational therapists rarely struggle because they do not know what to write. They struggle because the useful clinical detail is scattered across assessment notes, parent conversations, school emails, functional observations, goal reviews and the practitioner’s own memory. The referral letter then becomes another evening task: important, repetitive and surprisingly hard to start.

For Clawdelle, this is exactly the kind of Allied Health workflow worth improving first. Not diagnosis. Not automatic submission. Not pretending an AI system can replace clinical judgement. Just the admin layer around care: organising evidence, drafting a clear structure and giving the therapist something safe to review.

A good OT referral letter is not just a polite handover. It explains context, function, risk, goals and the reason another provider is being involved. When the letter is rushed, the next clinician receives fragments. When it is over-written, the OT loses time they should be spending on care, supervision or rest.

Quick answer

AI can help occupational therapists draft referral letters by turning existing notes into a structured first draft: reason for referral, relevant background, functional observations, current goals, risk considerations and requested next steps. The therapist still reviews, edits and signs off before anything is sent.

For most OT practices, the safest starting point is a human-in-the-loop drafting workflow connected to the notes and templates the practice already uses. That gives speed without handing over clinical responsibility.

Where referral letters slow OT practices down

OT referral letters often sit at the intersection of clinical reasoning and admin. They need enough detail to be useful, but they are usually written after a full day of sessions, travel, documentation and family communication.

The common bottlenecks are predictable:

None of this requires magic. It requires a calmer document workflow.

A safer AI workflow for OT referral letters

The wrong approach is “AI writes the letter and sends it”. That is not the move. The better workflow is draft-first, review-always:

  1. Capture the source material. Session notes, assessment summaries, goal progress, family concerns and relevant functional observations are collected in one place.
  2. Choose the audience. The practitioner selects whether the letter is for a GP, specialist, school, plan manager, support coordinator or another clinician.
  3. Generate a structured draft. The draft follows the clinic’s preferred format and tone.
  4. Highlight uncertainty. Missing details and assumptions are surfaced instead of hidden.
  5. Clinician reviews and signs off. The OT owns the final document.

That last step is non-negotiable. In Allied Health, speed is only useful if accountability stays clear.

What a good OT referral letter should include

A practical structure for many occupational therapy referrals looks like this:

SectionPurposeExample content
Reason for referralMake the request explicitAssessment, further investigation, shared care, school support, functional capacity input
Client contextOrient the recipient quicklyAge, setting, relevant diagnosis if already established, family or support context
Functional observationsShow what is happening day to daySelf-care, sensory processing, regulation, mobility, handwriting, play, participation, executive function
Current goalsConnect the referral to therapy prioritiesIndependence, participation, school access, home routines, emotional regulation
Risk or urgencyClarify why timing mattersEscalating distress, safety concerns, delayed access, family strain, school transition
Requested next stepPrevent vague handoversReview, assessment, report, care coordination, treatment recommendation

Clawdelle’s role is to help organise these ingredients. The therapist’s role is to decide what belongs, what needs softening and what requires clinical precision.

Worked example: from notes to referral draft

Imagine a paediatric OT has six months of notes for a child who is struggling with classroom participation, emotional regulation and morning transitions. The raw notes contain useful observations, but they are not letter-ready.

A draft-first AI workflow can turn those notes into:

The clinician then edits the draft: removes anything too speculative, adds context the AI could not know, checks consent, and ensures the letter matches the practice’s standards.

Common mistakes to avoid

Key takeaways

FAQ

Can AI write referral letters for occupational therapists?

It can draft them from existing notes and templates, but the occupational therapist should review, edit and approve the final letter before it is sent.

Is this safe for NDIS or paediatric OT work?

It can be safe when used as a drafting tool with human review, privacy controls and clear limits. It should not be used to invent evidence, make clinical decisions or submit documents automatically.

What practice software does this replace?

It does not need to replace practice management software. For MVP 1, Clawdelle is focused on drafting and organising admin outputs around the systems a practice already uses.

Who should approve the final letter?

The treating clinician or authorised practitioner. AI can help with structure and wording; it should not own the clinical responsibility.

Want clinical admin drafts without losing control?

Clawdelle helps Allied Health practices draft notes, referral letters, recalls and reports for human review.

Start setup