.webp)
AI-powered patient comms:
What, why and how?
Ellie Thomas, Director at Camino and specialist in rare disease comms, and Iona MacKillop, Medical Writer at Camino with an interest in patient-centricity.
We believe patient education deserves just as much attention as communication aimed at other audiences like healthcare professionals – and should never be an afterthought. This is particularly true in complex areas like rare diseases, where we’ve seen first hand how good communication can mean the difference between feeling empowered and feeling completely lost.
We're passionate about making sure patients get the information they need – clearly, quickly and in a way that feels relevant.
Budgets are tight, which is why we’ve also been exploring how AI might help us get there. It’s not perfect (more on that later) but when used in the right way, it presents an opportunity to transform how we develop communications for patients with complex or rare conditions.
Here’s a quick run-through of what we’ve learned: what AI can do, why it matters, and how it’s already making a difference.
What: AI-powered patient communication
In our experience, AI tools (like ChatGPT, Claude and others) can help turn dense scientific content into something that feels more human (ironically 🙂) – and crucially, easy to understand.
When it comes to rare disease, this matters. Not only because rare diseases are often extremely complex with convoluted diagnosis and treatment pathways, but also smaller patient populations and often tighter budgets mean there are fewer resources available for dedicated patient comms. AI offers a way to do more, more quickly, and without sacrificing clarity or accuracy.
Why: Patients shouldn’t be the last to know
Patients deserve timely, relevant information that speaks to them – not just to their doctor. But in reality, there are a few common obstacles:
- Delays in translating clinical data for patients – Often there’s a big gap between HCPs getting access to new information and patients seeing it in words they can understand.
- Overly complex language – Scientific content is often full of jargon that’s tough for even the most motivated patient to decode. Limited budget for patient materials – HCP content often gets the lion’s share of funding.
- Limited budget for patient materials – HCP content often gets the lion’s share of funding.
- One-size-fits-all content – Rare disease journeys are rarely linear, and content that doesn’t reflect the stage or needs of the individual can fall flat or cause undue worry.
- Time-consuming localisation – Adapting materials for different cultures and regions is essential, but not always easy to do quickly or well.
AI helps to tackle all of these – increasing speed, lowering costs, simplifying language and making it easier to adapt and personalise materials. It means we can move towards a world where patients are informed partners in their own care.
How: AI in action for patient-centred communication
There are five key ways we’ve used to AI support more effective patient content:
- Timely patient versions of content – AI enables us to produce plain-language summaries at speed, so patients get the information they need without delay.
- Simplified language – Tools using natural language processing (NLP) can break down complex terms into language that patients actually use and understand.
- Faster translation and adaptation – AI can provide first-draft translations that help us localise content more quickly and identify cultural nuances earlier in the process.
- Smarter use of budget – With some of the legwork automated, more budget can go toward design, delivery and patient engagement.
- Tailored messaging – AI can help us draft content suited to different patient profiles (e.g. newly diagnosed vs. in long-term management), making education feel more personal.
What happened when we tried it
We recently worked in a rare genetic disorder where educational resources were pretty limited, and what did exist wasn’t particularly accessible to patients. Our goal was to develop clear, culturally appropriate and personalised patient education. Here’s how it played out:
- Tool choice: We tested a few AI tools. Initially we tried ChatGPT, but found the tone a little too American for our UK-based audience. So we switched to Claude, which offered a more natural-sounding tone – better suited to the client and end users.
- Localisation: With Claude, we could also tweak language and references for different regions, ensuring the content reflected local care pathways.
- Speeding up the process: We used AI to scan treatment guidelines and published research, then create plain-language summaries. This gave us a fast, solid first draft to build on.
- Tailoring to the journey: We prompted the AI tool to adjust tone and content according to where someone was in their journey – newly diagnosed, managing symptoms, or exploring new treatment options.
- Results: Within a week, we had patient-friendly materials ready to go. They were accurate, accessible, and well received by both the client and the patient advocacy groups they worked with.
What happened when we tried it
AI is brilliant at speeding up the ‘science-to-story’ process. It helps simplify language, tailor messages and get content out faster. AI is surprisingly good at creating content that is empathetic and emotionally resonant. There are some interesting studies that show this as well, including one which showed that AI outperformed doctors in answering patient queries with empathetic responses.1 But this process also confirmed what we already knew – that AI always needs human oversight, especially when it comes to ensuring scientific accuracy.
While we can appreciate and benefit from all that AI has to offer, we can’t blindly rely on its output when we’re developing medical content. In this, and every project where we use AI, we make sure to combine the AI-generated content with thorough reviews for quality and accuracy to make sure we’re always telling the right story.
What we want to try next
In an ideal world, we’d always want to co-create patient communication materials with people who are actually affected by the condition we’re talking about. But in rare disease, where the patient populations (and funds for co-creation) can be small, it’s not always feasible to get patients or carers involved. And so, we have a burning question: Can AI act as the patient/carer focus group for us? If we create content, can AI critique and validate it from the patient perspective?
If so, it could open up yet another avenue for AI to help us overcome some of the barriers to effective patient communications and ensure we create the right content in the right format.
If this is something you’ve considered or are interested in exploring, we’d be keen to hear from you! We see a lot of potential for AI in the realm of patient communications…
What we want to try next
AI isn’t going to replace patient communications teams and their valuable expertise – but it is changing how we work. Used wisely, it’s a brilliant tool to speed things up, improve clarity, and expand what’s possible on tight timelines and budgets.
At Camino, we’re excited to keep exploring how AI can support more accessible, personalised patient education – especially in rare disease, where it’s needed most.
If you’re curious about the practical ways AI can transform communication across pharma teams, HCPs and patients, join us at Adventures in Pharma – our flagship conference dedicated to just that. Tickets are selling fast at adventuresinpharma.com – and you can use code FRIENDOFCAMINO for £200 off the list price.
References:
Harvard Health Publishing. Can AI answer medical questions better than your doctor? Available at: https://www.health.harvard.edu/blog/can-ai-answer-medical-questions-better-than-your-doctor-202403273028 (Accessed April 2025).
Want to know more? Get in touch!
