Tuesday 10 March 2026 | 15:15-16:15
Stream: Tech Innovation
Session format: Workshop
Chair: Jenny Shand UCL; England

Part 1 - Improving follow-up adherence among hypertensive and diabetes patients through mobile phone-based lifestyle coaching In rural Ethiopia: a randomized controlled trial

In many rural communities across Ethiopia, patients with chronic conditions like hypertension and diabetes face significant barriers to follow-up care—limited access to health services, low awareness of complications, and seasonal disruptions. In this session, I will share findings from my published research on self-care practices among hypertensive patients in North Shewa, and present plans for an upcoming nationwide randomized controlled trial. This trial will explore how mobile phone-based lifestyle coaching can improve medication adherence, encourage healthy behaviors, and ultimately reduce complications. I will also highlight lessons from my broader public health work, including outbreak response, health system strengthening, and sanitation initiatives. Delegates will gain insights into how low-cost, scalable interventions can improve chronic disease outcomes in resource-limited settings. This session is ideal for those interested in health equity, population health, and digital innovation in care delivery. 

After this session, participants will be able to:

1. Describe key barriers to chronic disease follow-up care (e.g., for hypertension and diabetes) in rural, low-resource settings like Ethiopia.
2. Explain how mobile phone-based lifestyle coaching can improve medication adherence, promote healthy behaviors, and support self-care among patients with chronic conditions.
3. Identify practical strategies for designing and implementing scalable, community-based interventions to strengthen chronic care management in underserved populations.

Hailemelekot Kebede Shewarobit Federal Prison; Ethiopia

Part 2 - A digitally enabled data-driven virtual ward round improvement intervention for preventing in-hospital harm to people with diabetes

In this session, we show how we combined AI with improvement methodology to develop virtual ward rounds—an approach that enhances care quality and safety by remotely screening and triaging people in hospital with diabetes using a data-driven clinical decision support system (CDSS).

People with diabetes are 25% more likely to be hospitalised, occupying around 20% of hospital beds—nearly three times their population prevalence. Most are admitted for reasons unrelated to diabetes, but face more complications and poorer outcomes.

Virtual ward rounds collate real-time data about all hospitalised people with diabetes, enabling earlier, proactive intervention by non-ward-based diabetes specialists to prevent harm and improve experience. This contrasts with traditional virtual wards, which focus on admission avoidance and early discharge, and CDSSs that typically deliver reactive alerts to non-specialists. We demonstrate how routinely collected data can reliably and transparently support real-time in-hospital case-finding and triage using explainable, AI-driven predictive models.

After this session, participants will be able to:

• Understand how to leverage AI and routinely collected data to improve real-time identification of people in hospital with diabetes at risk of harm.
• Learn how data-driven predictive models can effectively detect those at risk.
• Explore how to present model outputs in clear, accessible, and explainable formats that support quality improvement.
  

James Beveridge Imperial College London; England