A Personalised Early Warning Decision Support System with novel Saliva Bio-Profiling to Predict and Prevent Acute Exacerbations of Chronic Obstructive Pulmonary Disease - 'Predict&Prevent AECOPD'

Research question: Can a personalised early warning decision support system with built-in composite predictive algorithms be used by patients with COPD at home to improve self-management and help them predict exacerbations early, intervene promptly and avoid hospitalisation? Background: Patients are encouraged to recognise and treat acute exacerbations of COPD (AECOPD) but in many cases because of day-to-day variability in symptoms, events go unrecognised and untreated; this can lead to hospital admissions and long term decline. In addition, heterogeneity amongst COPD patients has hampered personalised AECOPD recognition to date. Our proprietary Early Warning Decision Support System (Predict&Prevent AECOPD) uses longitudinal home monitoring of relevant subjective and objective data (symptoms, spirometry, biomarkers) in real-time using a smart App connected to Bluetooth-enabled sensor peripherals. These data are used to construct COPD-relevant individual profiles; AI-driven algorithms can then identify changes in health status and imminent AECOPD to provide timely individualised alerts and guidance/action plans for patients. Aim and Objectives: Our overall ambition is clinical validation and commercialisation of the Early Warning Decision Support System to guide and support COPD patients in identifying exacerbations early, leading to a 25% reduction in total AECOPD-induced hospital admissions. Using a multi-skills team approach, over 36 months we will establish: Clinical- and cost- effectiveness of Predict&Prevent AECOPD in a community setting. Regulatory approval for the two system versions. Stakeholder engagement and an implementation plan. A robust business model and commercial road-map. Methods: Clinical effectiveness will be tested in a randomised controlled trial (RCT) of Predict&Prevent AECOPD conducted in 384 COPD patients who have had ≥2 AECOPD or ≥1 hospital admission for AECOPD in previous year. This is powered to detect a 25% reduction in AECOPD hospitalisations and will validate our exacerbation prediction models. As models are constructed to work using both blood and salivary biomarker measurements, in this study the target biomarker, CRP will be measured in both biofluids, in anticipation of salivary CRP analysers being on the market within next 2 years. This addresses end-user preferences for saliva sampling, preferred over blood/urine. Trial findings will therefore assist in validating salivary CRP analysis. The RCT data will also be used to develop an economic model to describe cost-effectiveness, which in turn will refine our current budget impact model for the NHS. Parallel work packages will assess: Acceptability of the System to patients through qualitative research with 30 participants Utility of the System to policymakers, through inquiry and stakeholder events Bottom-up market analysis for the product by our business partner. Overall, the results will enable production of a coherent value proposition and implementation plan, product CE marking and value added partnerships to assist market penetration and uptake within the NHS/other sectors. Anticipated impact and dissemination: Predict&Prevent AECOPD with its smart algorithms offers an efficient and precise solution to supporting COPD patients at home and preventing exacerbation-induced hospitalisations. Impact on COPD care pathway is anticipated through NHS adoption and uptake/commissioning partnerships and changes to national AECOPD guidance. Dissemination will occur through conferences, publications, marketing and via our PPI groups and newsletter.