Information Leaflets vs Artificial Intelligence: Comparing Perceptions of Stroke Survivors and Professionals in a Mixed-methods Study
This mixed-methods study reveals that while stroke survivors and researchers generally rate AI-generated stroke information positively and find it comparable to traditional third-sector leaflets, clinicians remain more skeptical, highlighting the need to prioritize the perspectives of those with lived experience in developing future stroke resources.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Imagine you've just had a major health event, like a stroke. You're leaving the hospital, and your brain is full of questions: What do I eat? Will I drive again? How long until I feel normal? You reach for a pamphlet or a website, hoping for clear answers. But often, those answers feel like they were written in a secret code only doctors understand, or worse, they aren't there at all.
This study is like a taste test to see what kind of "information meal" stroke survivors actually want to eat. The researchers compared two different chefs:
- The Traditional Chef: Human experts who write information leaflets for UK charity websites (the "Third Sector").
- The New Robot Chef: An Artificial Intelligence (AI) called ChatGPT, which was asked to write answers to the same questions.
Here is what they found, broken down simply:
The Setup: A Blind Taste Test
The researchers picked 15 of the most common questions people ask after a stroke. They gathered answers from the charity websites and asked the AI to write answers too. Then, they invited three groups of people to be the "food critics":
- The Survivors: People who have had a stroke (and their families).
- The Doctors: Clinicians who treat stroke patients.
- The Researchers: Scientists who study stroke.
The critics were given two answers for each question but weren't told which one was written by a human and which was by the robot. They had to guess the source and pick their favorite.
The Results: Who Liked the Robot?
1. The Doctors were Skeptical
The doctors were like strict food inspectors. They preferred the human-written leaflets. They felt the robot's answers lacked "soul" or empathy. Even though the robot gave good facts, the doctors thought the human version felt warmer and more trustworthy. They were the least likely to choose the AI answer.
2. The Researchers were Open-Minded
The scientists were more willing to try the robot chef. They liked the AI answers about 54% of the time. They seemed to appreciate the clarity and structure the AI provided.
3. The Survivors were Split (But Open to AI)
This is the most interesting part. The people who actually lived through a stroke didn't have a strong bias against the robot. In fact, they preferred the AI answers nearly half the time (49%).
- The Surprise: Even though the doctors thought the robot sounded cold, the survivors often didn't notice the difference. When they rated how "empathetic" (caring) the answers were, survivors gave the AI and the humans equal scores.
- The Catch: The survivors did notice when the answers were too long, too confusing, or sounded like a strict schoolteacher. They wanted answers that were short, clear, and gentle.
What Did the Survivors Actually Want?
Through their interviews, the survivors described their ideal information "recipe" using three main ingredients:
- Content (The Ingredients): They wanted the answer to go straight to the point. No wandering around in circles. If they asked about side effects, they wanted the facts, but explained in a way that made sense. If they asked about life after stroke, they wanted it short and focused.
- Structure (The Plating): Imagine a messy pile of spaghetti versus a neatly arranged plate. Survivors hated the "messy" long paragraphs. They loved bullet points and short sentences. One survivor mentioned that after a stroke, their memory isn't as sharp, so long blocks of text are hard to digest. They wanted the information "plated" in a way that was easy to pick up and eat.
- Tone (The Seasoning): This was crucial. They didn't want the answer to sound like a cold medical textbook or a scary lecture. They wanted it to sound like a friend talking to them—warm, gentle, and understanding. However, they didn't want the "friend" to be too emotional when giving practical advice (like how to get financial help); they just wanted it to be kind and clear.
The Big Takeaway
The study found that the current "menu" of stroke information is incomplete. Some questions simply don't have answers on the charity websites.
While doctors are worried about the robot chef, the people who actually need the food (the survivors) are saying, "Hey, the robot's food is actually pretty good, as long as it's not too complicated."
The paper concludes that if we want to help stroke survivors, we shouldn't just ask the doctors what they think is best. We need to listen to the survivors themselves. They are telling us that whether the answer comes from a human or a robot doesn't matter as much as how it is written: it needs to be clear, short, kind, and easy to understand.
In short: The robot isn't trying to replace the human doctor, but it might be a very helpful assistant for writing information that survivors can actually read and understand.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.