Talking about dying is hard. It's hard for patients, families, and doctors. We know we need to get better at these talks, especially in end-of-life care. Old ways of training doctors often don't help enough.
Two brothers sought to change that (per The Independent). They used AI, the study of language, and advice from care experts. They want to make these tough talks better for everyone. This idea fits with a wider push for doctors to use simpler words.
The linguistic research on palliative care
Dr Bob Gramling is a doctor who helps people near the end of life. He learned how much words matter when his own father was dying. This was true even though Bob was a doctor himself. He said it showed him "how vulnerable people" are without medical knowledge. So, he started to study how people talk about dying. He created the Vermont Conversation Lab. It holds recordings of talks with hundreds of patients near death. That's over 1.2 million words.
What's the goal? To find parts of talks that help patients feel heard. Bob works with his brother, David Gramling. David studies language. He looks closely at how talks work, like who speaks when. He also studies why communication sometimes fails.
Their 2019 book, Palliative Care Conversations, shows how talks work. Sometimes, doctors break the usual rules of chatting and connect better. Other times, small talk fails, and people don't connect. Bob notes that doctors often talk a lot. But in end-of-life care, talks need more back-and-forth. "This is not a clean, logical experience... it’s about people connecting," Bob says.
The research also shows that patients face big challenges. David calls this "language in extremis." Pain, tiredness, and medicine can make it hard to talk. Patients might struggle to find words or understand things. They may not have the energy for careful talk. But they still try to follow social rules, like being polite.
Where readability comes in
This imbalance means doctors have a big duty. They need to listen better. They also need to use much clearer language. This need for plain speaking is being seen across healthcare. The NHS in the UK told doctors to write letters patients can understand. This is part of the 'Please Write to Me' plan, backed by top medical groups (‘NHS doctors told to use plain English’).
It means using simple words. For example, say "twice daily" not "bd" (a Latin term). Say "breathlessness" not "dyspnoea." Say "brain" not "cerebral." Use short sentences. Avoid jargon and letter codes (acronyms). Dr Hugh Rayner, who leads the plan, said writing clearly is crucial. If plain English is key for regular letters, it's even more vital in end-of-life care, like the Gramlings study.
Patients are dealing with serious illness and worry. Doctors must share news with care and clarity. Just asking a patient to "speak openly" doesn't work if the words are too complex. So, readability is essential to making things easier. It works alongside the better listening the Gramlings talk about.
More than words
Bob's team also uses computers to study silences in talks. A helper, Brigitte Durieux, listened to many pauses. Sometimes the talks made her cry. The team used AI to find pauses longer than two seconds. Then people checked if these pauses showed emotion. They found some "connectional silences." These were short, rare pauses (under 4 seconds). They seemed to mark moments of real feeling or understanding.
After these silences, talks often changed. Patients might talk more. They might lead the chat. A shared quiet moment can show shared feelings. It can lead to more understanding. Kate's mother had a key end-of-life talk with Bob. Kate said, "Where there’s silence... that’s where the caring shows up."
Experts agree that listening is hard for doctors. Wen-Ying Sylvia Chou, who funds research, says doctors aren't often trained to just listen. The Gramlings think AI can measure these quiet parts of talks. This could help hospitals reward doctors for having better talks.
Other research
Other scientists are also looking at AI for this. James Tulsky in Boston works on computer programs to spot empathy in talks. The technology is growing. But there are still things to figure out. The Gramlings' work so far is only with English-speaking cancer patients. Talks might be different in other cultures or with other illnesses.
Some people doubt that computers belong in these personal moments. Bob understands this worry. He was afraid of making a "sacred experience" into simple, maybe "meaningless" data. To respect the people involved, his lab has a small ceremony. They read a patient's name from the study and ring a small bowl.
In the end, the Gramlings say that studying the 'science' of talks helps the 'art.' Learning about language, pauses, and turn-taking gives tools to improve. Putting this knowledge together with the need for plain English (readability) everywhere in healthcare is key. It can help doctors do more than just give facts. It can help them have real talks where patients feel truly understood at the end of life. As Kate said, maybe new tools, used with care and clarity, can make these vital human moments "more wonderful."
Source: The Independent
