Rheumatoid arthritis (RA) is the most common inflammatory arthritis affecting 1% of the population. In recent years the outcome for patients with RA has improved significantly although it remains a chronic rather than curable disease resulting in a health economic burden to the UK of £8Billion/annum. Good outcomes for patients are associated with tight control of disease measured through a composite disease activity score (DAS), which integrates swollen and tender joint counts, patient visual assessment score of disease activity with markers of inflammation (CRP or ESR). However tight disease control, and good outcomes for patients, requires intensive DAS monitoring in order to ensure timely adaptation of therapy, a situation that is frequently not possible due to limited health care resources. Remote DAS monitoring e.g. using a mobile app offers the opportunity to deliver frequent disease activity measures whilst also engaging patients in disease management, an approach recommended by the European League Against Rheumatism and also preferred by a majority of patients. With this aim, Barts Charity recently funded an 18 month pilot study (completion due December 2018) to evaluate the impact on health service delivery of a bespoke mobile app for DAS (BioT-APP) developed within Experimental Medicine and Rheumatology. However although mobile-apps are capable of delivering disease activity metrics they are currently not able to support patient/clinician decision making to optimize therapy, and therefore require frequent clinician review of outputs. Integration of artificial intelligence (AI) to an app has the potential to offer remote disease activity monitoring with guided “intelligent” care plans offering timely therapy optimization and improving overall health care outcomes.

PAMBAYESIAN will create AI prototypes using Bayesian networks to support both clinicians and patients remotely. To model the AI prototypes, clinicians will provide medical expertise including analysis of clinical decision making processes, background knowledge, and interpretation of clinical data. The project intends to use qualitative techniques to examine the acceptability of AI prototype to patients through patient groups and rheumatology clinicians. The AI prototype(s) will be also integrated into the BioT-APP to develop the intelligence needed in this mobile app, and benefit the platform of BioT-APP to facilitate the interaction between clinicians and patients. The AI prototypes of PAMBAYESIAN will support both clinicians and patients remotely, and this will realize the tight control of RA disease and cause to reduce the economic burden on the healthcare system.