A machine-learning system will predict COPD flare-ups days before they happen, then trigger interventions to keep patients out of hospital. COPD is a progressive lung disease with no cure. Its hallmark is sudden exacerbations—acute worsening of symptoms that often require emergency hospital care. These events accelerate lung decline and reduce quality of life. The researchers’ pilot study showed that AI can forecast these exacerbations several days in advance, creating a window to act. The problem is that no system yet exists to turn those predictions into reliable, clinically safe, and patient-accepted interventions. If this succeeds, patients could receive early warnings and tailored advice—such as adjusting medication or avoiding triggers—before symptoms spiral. That would reduce both the frequency and severity of exacerbations, slowing lung function decline and improving daily life. For the NHS, fewer emergency admissions and better-targeted use of medical resources would lower acute care costs. The project also aims to build a regulatory-approved digital platform that could be adopted across the health service, creating a competitive UK digital health ecosystem for chronic disease management.
View original technical description
Background : Chronic obstructive pulmonary disease (COPD) is a common disease and is predicted to become the 3rd leading cause of death by 2030. Because there is no cure, treatment involves long-term management of symptoms to slow the decline of patient health and improve quality of life. A key characteristic of COPD is the presence of ‘exacerbation events’ defined as an acute worsening of symptoms. Exacerbation events are detrimental to health and often lead to hospitalization. Avoiding exacerbation events is desirable because it leads to higher quality of and longer life. Our recent pilot demonstrated that machine learning can predict acute exacerbation events several days in advance. Such predictions have huge potential to improve clinical models and drive behavioural change necessary to avoid acute exacerbation. Aims : We aim to predict COPD exacerbation events and use them to trigger interventions. The objectives are to 1) develop safe and reliable AI, 2) ensure acceptance and engagement with patients and clinicians, 3) establish clinical efficacy, safety, regulatory data, 4) establish a competitive AI-enhanced digital health ecosystem. Workplan : Four work packages will address clinical, technical and adoption. WP1 “Person-Centred Care” will create patient journey and clinical model, and work with clinicians, patients, and the public to ensure effectiveness and safety. WP2 “Agile Digital Platform Engineering” will develop a digital platform that incorporates AI components from WP3 “Predictive Analytics and Visualisation”. Finally, WP4 “Dissemination, Certification and NHS Adoption” will engage stakeholders, obtain regulatory approvals, and secure investment. Impact: By predicting and mitigating exacerbations, we aim to empower patients to engage with the management of their COPD and reduce the burden on NHS services. Exacerbation frequency/severity is expected to be reduced, slowing deterioration of lung function, and increasing quality of life. ROI is anticipated from reduced use of acute care and better use of medical interventions.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know