Background and Aims

Prevalence of type 2 Diabetes in Mexico is high. Early detection and intervention of the condition, particularly in children and adolescents will have a great impact in the wellbeing of individuals and reduce the cost of treatment. CARDIGAN aims to test whether Criticality Analysis (CA) can be applied to gait data to give a reliable and cost-effective way to detect individuals at risk of developing type 2 diabetes.

Methods

Gait data and clinical data of overweight children was collected using portable Inertial Measurement Units in a 6-week trial, conducted by Hospital Infantil de Mexico Federico Gomez. Data analysis is now being performed in a semi-blind manner using a novel machine learning approach based on CA.

Results

The results obtained from the CA show visible differences in the gait patterns among the obese participants, and even more pronounced differences are seen compared with the control group. Data from each week will be analysed to track the progression of the participants. All of these results will be cross-referenced with the clinical data obtained.

Conclusions

The use of Criticality Analysis of gait as a means of diabetes assessment is evaluated in order to track the progression of the participants’ condition, and detect if they are at risk of developing type 2 diabetes. This will aid in determining the appropriate course of intervention.

This work was supported by an Institutional Links grant, 432368181. The grant is funded by the UK Department for Business, Energy and Industrial Strategy and delivered by the British Council. For further information, please visit www.newtonfund.ac.uk.

Background and Aims

PEPPER (Patient Empowerment through Predictive PERsonalised decision support) is an EU-funded research project which aims to improve self-management of type 1 diabetes (T1D). The system comprises an AI insulin bolus recommender, coupled with a safety system. The aim of the qualitative arm of this clinical feasibility study was to examine the context of participants’ interaction with the PEPPER system and identify incidents where bolus recommendations were trusted and accepted.

Methods

This was a multicentre (UK and Spain) non-randomised open-labelled 6-week pilot study. Thirteen adults with T1D participated in weekly telephone interviews to explore the context of their interactions and responses to PEPPER. Data was thematically analysed through conceptual frameworks for engagement with healthcare digital behaviour change interventions.

Results

Participants reported their key interactions as responding to PEPPER bolus recommendations, inputting carbohydrate values, interpreting continuous glucose monitoring (CGM) values through visualization of personal data and dealing with safety alarms. Two themes were associated with trust and engagement with the system; ‘feeling monitored’ and ‘feeling in control’. The incidents where participants trusted PEPPER also enhanced personal expertise of T1D through insights provided by the safety system such as low glucose basal insulin for pump users. Benefits were balanced against technical challenges of the system, which were used to improve the PEPPER application and enhance user experience.

Conclusions

Some participants suggested that even access to PEPPER for a temporary period could positively influence self-management strategies. Contextual interviewing is a valuable tool in mobile application development for diabetes decision support systems.