Drafting Post-MDT Patient Letters

The problem

After a multidisciplinary team (MDT) meeting, you have a decision. Now you need to write to the patient. The clinical content is clear. The difficulty is translating it — staging, treatment modality, proposed sequencing — into a letter that a frightened person sitting at their kitchen table can actually follow.

These letters matter enormously. Patients re-read them. Families read them. They are the written record of what was decided and why. Writing them well takes time and care that is hard to find at the end of a full clinic day.

How AI helps

An AI tool can take a set of clinical bullet points — the key facts from the MDT — and produce a structured, plain-language draft letter. You provide the clinical content. The tool provides the first version of the words.

This is not the tool making clinical judgements. It is the tool doing the structural and linguistic work: organising bullet points into readable paragraphs, choosing plain language over clinical jargon, and maintaining a calm, warm register throughout. The clinical accuracy, the tone calibration for this patient, and the final sign-off are entirely yours.

The tool works well for: explaining the rationale behind a treatment decision, summarising what will happen next and when, and setting out what the patient should watch for and who to contact.

A real example

Dr Kwame is a consultant medical oncologist. His MDT has agreed a plan for a patient in his mid-sixties with newly diagnosed colorectal cancer — surgery followed by adjuvant chemotherapy. The patient has asked for everything in writing. He has no medical background and his daughter, who is his main support, lives abroad and reads all his letters.

Dr Kwame opens Claude and types:

Draft a patient letter explaining the following MDT decision for a man 
in his mid-sixties with newly diagnosed colorectal cancer.

MDT decision:
- Surgery recommended as first step
- Adjuvant chemotherapy to follow surgery pending pathology
- Oncology review appointment to be scheduled post-surgery
- Patient to be seen by clinical nurse specialist before surgery

The patient has no medical background. His daughter reads all his letters 
and may not speak to his doctor directly. Write clearly and warmly. 
Explain each step in plain English. Do not include specific drug names, 
doses, or prognosis. Maximum 350 words.

The tool produces a draft. Dr Kwame reads it carefully. He adds a sentence acknowledging that receiving this news is difficult. He removes a phrase about "excellent outcomes" that the tool included — which he does not wish to imply. He adjusts the description of surgery to reflect his unit's specific approach. The letter is ready in eight minutes.

Try it yourself

Things to watch for

The tool tends toward optimism. AI tools often add reassuring phrases — "your care team is here to support you," "this plan gives the best chance of success." Some of these are appropriate. Some are not, for some patients. Read every sentence for tone as well as accuracy.

It does not know the patient. The letter it drafts is for a generic patient. You know whether this patient prefers directness, needs more reassurance, or has specific anxieties the letter should acknowledge. That calibration is yours.

It cannot include your local contact details, pathways, or dates. You will need to add these yourself. Treat the draft as a framework, not a finished product.

Plain language is not always precise language. Simplification sometimes creates ambiguity. Read the clinical content as you would read any letter before it goes to a patient under your care.

Remember: AI is a helpful assistant, not a clinician. You make the call.