April 10, 2018

Digital communication between patient and clinical team forms basis for Artificial Intelligence

Writing about web page https://warwick.ac.uk/fac/med/research/hscience/sssh/research/lyncs/

People with a health condition have long stretches of time between encounters with their healthcare team when they get on with living with their condition. People experience change in their condition, treatment effects and side effects mostly on their own - without engagement with their healthcare team. They have to interpret what they experience themselves. One of the most memorable comments a patient made to me as a young doctor was:

Going home with a prescription for a new treatment can feel very lonley.

Accessible information and explination makes a difference, but even where we have good evidence about a condition and its treatment we still only know what usuallyy happens most of the time. We cannot predict exactly what will happen to one particular patient1.

In our study of young people living with longterm conditions (the LYNC study)2 it was at times of change - new treatments or worsening symptoms or new life experiences such as going to univeristy - that young people most appreciated and benefited from digital access to their healthcare team. They were able to text, email or phone about what they were experiencing and receive interpretation, reassurance and guidance.

Communication via digital channels is easily recorded and stored. By enabling patients, between routine appointments, to digitally contact their clinical team we can build datasets of their concerns and the team's responses. This can be used as the training dataset for a chatbot based on artificial intelligence (AI), working alongside the clinical team. Initially the chatbot might have sufficient data to learn how to respond to common questions, for example, how to take a medication. As the dataset grows, the AI algorithm can learn how to respond to more complex clinical questions. By linking with patient records, responses can be tailored to individual patients. The AI chatbot can also learn from what happens to patients and does not forget. In theory an AI chatbot could become as good or even better than a clinician, although there would still remain uncertainty.

Clinician and chatbot working together could be quite a team and counter each other's cognitive biases.

This type of development can begin right now but there are ethical and social implications that need our attention.

1. Gorovitz S and McIntyre A. Towards a Theory of Medical Falibility. The Journal of Medicine and Philosophy. 1976:1(1):51-71

2. Griffiths F, Bryce C, Cave J et al. Timely digital patient-clinician communication in specialist clincal services for young people: a mixed-methods study (The LYNC Study). Journal of Medical Internet Research. 2017;19(4):e102

- No comments Not publicly viewable

Add a comment

You are not allowed to comment on this entry as it has restricted commenting permissions.


April 2018

Mo Tu We Th Fr Sa Su
Mar |  Today  | May
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29

Search this blog


Blog archive

Not signed in
Sign in

Powered by BlogBuilder