How to Use Claude as an Allied Health Assistant: The Lucas Skill Guide

Allied health professionals — physiotherapists, occupational therapists, speech and language therapists, dietitians — spend a striking share of the day not with patients but on documentation: assessments, treatment plans, progress notes, referral letters. Used carefully, Claude for allied health can take the drafting burden off that load. An AI allied health assistant is a tool for qualified practitioners, not a source of clinical advice.

Lucas is that tool. He is a documentation persona you load once into Claude, ChatGPT, or any AI chat, and he is built around two principles that keep clinical writing safe: he works in the structure of your discipline, and he never invents clinical detail. Where a finding or measurement belongs, he leaves a clearly marked placeholder for you to complete. Every output is a draft for the clinician to review and sign.

Why generic AI clinical documentation is risky

An unguided model will fabricate clinical detail — ranges of movement, scores, history — that was never provided, which in a health record is dangerous rather than merely sloppy. It uses the wrong structure for the discipline, blurs the line between disciplines, and implies a draft is ready to file when it is not. In clinical documentation, plausible-but-invented is the worst possible output.

What changes with the Lucas skill

Lucas asks which discipline and which document type before he starts, then drafts in the right format — SOAP notes, an assessment, a care plan, a referral. He uses [PATIENT] and [DATA] placeholders rather than inventing findings, flags what needs the clinician’s completion or judgement, and keeps every output explicitly a draft. The point is to remove the typing and structuring, never the clinical reasoning.

What it actually produces

Assessment frameworks, treatment and care-plan drafts, progress-note structures, referral and discharge letters, SMART goal-setting frameworks, and patient-facing instruction drafts — each in the conventions of the relevant discipline, with placeholders where clinical detail is required.

How to get the most out of it

State the discipline and the document type up front. Give it the real clinical detail you want included — it will not invent it, so the draft is only as complete as your input. Treat every output as a first draft, and have the responsible clinician review, complete, and sign it before it enters a record.

Who this is for

Qualified allied health professionals, private practices, and clinics carrying a heavy documentation load. It works with Claude, ChatGPT, or any AI chat that accepts a system prompt. For related clinical writing, the health & medical skills collection covers consent, clinical summaries, and patient materials — each a focused assistant rather than a general chatbot. Lucas is a documentation tool for qualified clinicians; its outputs require clinician review and are not medical advice.