ESPE Abstracts

Aim of the study: In type 1 diabetes, it is well recognized that collecting additional information about diet, physical activity, health status, stress and any patients’ everyday behavior, is crucial to evaluate accurately metabolic control and therapeutic prescription adherence. The aim of this study is to test AID-GM (Advanced Intelligent Distant – Glucose Monitoring) a web-based platform, able of collecting automatically patient generated health data (PHGD) coming from different sources (blood glucose sensors, activity trackers, vocal messages, etc).

Methods: Thirty young TDM1 patients (over 11 years), under multiple daily injection therapy and using a FGM sensor (FreeStyle Libre^®, Abbott Diabetes Care, Alameda, CA) will be overall enrolled. To determine time spent walking and sedentary time, a Fitbit device will be delivered. AID-GM will be used to automatically collect and share data coming from these sensors and provide several advanced analysis and visualization tools. Moreover, a mobile app will be used by patients to record vocal messages reporting any relevant health information. An automatic tool will extract the information from messages and store them into the system database. Finally, the system is able to automatically detect specific temporal patterns in single or group of patients’ data, like for example rebound effect and dawn effect. The temporal analysis can be focused on specific time frames (e.g. days of the week, moments of the day, etc).

Results: The application is easy to use by both patients and care providers. It offers several functionalities such as: 1) rapid integration of clinical and health related data and sharing between patients and physicians, without having to use paper-based diaries; 2) active on-line inspection and analysis of real-time generated data for health status monitoring and prevention/prediction; 3) active monitoring of treatment outcomes (even remotely); 4) health-related data access. Finally, a potential usage of the platform will be in the context of clinical research trials running in realistic day-by-day settings.

Conclusions: The platform supports effectively home care supplying every information and analysis tools useful to increase knowledge about the factors influencing the patient’s glucose metabolic control. Using the platform, it will also be feasible to design observational clinical trials collecting PHGD at low cost with long follow-up with the aim of deriving model-based indexes of glucose metabolism and increasing the insight on basic mechanisms underlying diabetes disease.