Is ROMI, a conversational artificial intelligence, an effective way to deliver support and structured education to people with type 2 diabetes who do not engage with usual care services?
What is the research question?
Can virtual artificial intelligence (AI) diabetes team member improve patient outcomes and NHS diabetes team reach and efficacy?
What is the problem?
The NHS Digital National Diabetes Audit in 2018 reports that around 3.9 million people in the UK are affected by Type 2 Diabetes (T2D), often impacting negatively on individual quality of life. Six in every ten people with T2D are not meeting their recommended treatment targets that will reduce their risk of complications. Two of our study team, who run NHS diabetes teams, know this from their daily struggle to meet patient needs.
What is the aim of the research?
Our aim is to build a working virtual artificial intelligence (AI) diabetes team member called ROMI (Relational Online Motivational Instrument), both to show that it can be accepted by patients and staff, and to show that it works in promoting self-management. People with T2D will communicate with ROMI via a phone/tablet/PC, like with Siri, but ROMI also has an animated avatar on the screen. Through conversation, ROMI offers information and acts as a mentor promoting positive lifestyle changes. Over time, ROMI learns what helps and hinders the patient, with the aim of improving their self-management. We will recruit people with T2D from a wide range of communities to use ROMI. We will ask them if they think that such a system would improve their self-management of T2D. If so, then ROMI will improve outcomes and provide NHS teams with valuable patient information so they can improve care. We also want to see if ROMI can reach patients that currently do not use NHS services.
How will this be achieved?
After PPI work examining acceptability, ROMI will be developed and tested by 40 patients with T2D, recruited via community groups. It will then be refined and tested by another 30 patients. Usage data collected will be analysed alongside user and healthcare professional opinions about the function and acceptability of ROMI.
Who is leading the research?
Rik Lander, Senior Lecturer, Faculty of Arts, Creative Industries, and Education, University of the West of England.
Further information:
For more information or to get involved in this project, please contact bnssg.research@nhs.net.