All 2 entries tagged Lync Study

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

Frances Griffiths : 10 Apr 2018 17:53 |  Tags: Artificial Clinical Communication Digital Technologies Intelligence Lync Study |  Comments (0) |  Close comments |  Report a problem

Where to start with digital communication with patients?

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

Design a bespoke platform for your clinical service or use existing digital systems?

One answer is to do both! Evidence from our empirical study [1] suggests that when introducing digital communication with patients, changing the working patterns of the clinical team is the difficult part. There needs to be adaptation and flexibility, both on the side of the clinical team and on the side of the new intervention - the digital communication system.[2] The ideal might be, for example, a bespoke patient portal. By introducing some elements of the patient portal using existing digital systems gives time for change in work patterns. It also provides insights for the design of the portal. Patients interviewed in the LYNC study continued to use digital communication channels with their clinical team if they were confident the clinical team would respond.[1] So, if we want patients to use a patient portal the clinical team has to be responsive from the moment it is launched.

Consider this example. An oncologist wants to develop a patient portal that can

• Store for patient viewing, the patients individualised treatment plan and any updates
• Provide information for patients to read about their treatment including when to contact the clinical team
• Send reminders for patients receiving chemotherapy for blood tests, chemo sessions and review appointments
• Enable patients to view their blood test results and analysis of trends
• Enable the patient to contact the clinical team when they have symptoms of concern.

At the back end of the patient portal, the clinicians want to:

• View the patients clinical notes while on the move to inform their response to patients contacting them
• Enter information about patient encounters in the clinical notes while on the move.

The oncologist has already identified potential savings from this innovation including reduction in wastage of chemotherapy drugs and reduction in number of A&E attendances.

Oncology patients have a lot to deal with, so a patient portal where everything they need is in one place seems ideal – although we need to check this with patients. Not everyone in the clinical team is as engaged with the idea as our oncologist but there are some potential quick wins that might persuade them. For example, text message reminders improve patient attendance at clinic appointments.[3] The hospital already has a reminder system for standard outpatient appointment. Could the existing system be used to send reminders for blood tests and non-routine appointments?

Testing out the use of video-conferencing using the existing hospital system would allow the enthusiasts in the oncology team to firstly work out how to integrate it into their work pattern and secondly, to assess how much advantage it provides over telephone – is it a necessary part of the patient platform? In the LYNC study we undertook a scoping review of videoconferencing.[1)] Patients mostly like it and sometimes it improves health outcome, but it does not usually save clinician time or save money. However, where there is a specific reason for using it, for example saving travel time for appointments, it can be of benefit. In our example the availability of videoconferencing might provide more information about a patient’s condition than telephone, so the clinician might be able to reduce the number of patients making a trip to the hospital.

Evidence from the LYNC study [4] suggests patients benefit from being able to engage digitally (email, text, mobile phone) with their clinical team when they most need it. This service can be offered using existing systems – hospital email and mobile phones. What is difficult is working out who is going to respond to patient contacts, when and where. Come the launch of the patient portal any new work patterns need to be well established. Our LYNC study Quick Reference Guides [5] suggest what needs to be thought through when offering patients digital engagement with their clinical team.

All innovation needs resource for its implementation. I have suggested digital engagement with patients is introduced in small steps. A danger is that the resources needed for implementation of the small steps is ignored so the innovations fail or the enthusiasts bear the burden. When a new digital system is introduced in one go, the resource needed for implementation is less easy to ignore.

-- Frances Griffiths, Professor of Medicine in Society and Penny Kechagioglou, the Oncologist

References:

1. Griffiths F, Armoiry X, Atherton H, Bryce C, Buckle A, Cave J, et al. The role of digital communication in patient-clinician communication for NHS providers of specialist clinical services for young people (The LYNC study): a mixed methods study. NIHR Journals Library Publications, HS&DR. 2018 in press.

2. May CR, Johnson M, Finch T. Implementation, context and complexity. Implementation Science. 2016;11(1):141.

3. Guy R, Hocking J, Wand H, Stott S, Ali H, Kaldor J. How Effective Are Short Message Service Reminders at Increasing Clinic Attendance? A Meta-Analysis and Systematic Review. Health Services Research. 2012;47(2):614-32.

4. Griffiths F, Bryce C, Cave J, Dritsaki M, Fraser J, Hamilton K, et al. Timely digital patient-clinician communication in specialist clinical services for young people: a mixed-methods study (the LYNC study). Journal of Medical Internet Research. 2017;19(4).

5. LYNC study team. LYNC Study Quick Reference e-book and Topic Guides: University of Warwick; 2017 [10th June 2017]. Available from: https://warwick.ac.uk/fac/med/research/hscience/sssh/research/lyncs/outputs/

Bronwyn Harris : 02 Feb 2018 16:40 |  Tags: Clinical Communication Digital Technologies Lync Study |  Comments (0) |  Close comments |  Report a problem