Towards a mobile solution for predicting illness in Type 1 Diabetes Mellitus: Development of a prediction model for detecting risk of illness in Type 1 Diabetes prior to symptom onset

Abstract

Illness in Type 1 Diabetes Mellitus (T1DM) patients makes it complicated to perform sufficient self-care, resulting in prolonged episodes of hyperglycemia and fluctuating blood glucose (BG) concentrations. Prolonged episodes of hyperglycemia elevate the risk of the patient developing diabetic complications, which makes infections such as common cold, influenza and influenza like illness more harmful for T1DM patients than the normal population. TTL, NST and AAU are researching a method of predicting illness in T1DM patients, using patient observable parameters. Daily BG measurements are identified as a relevant patient observable parameter, due to early rise when infected and elevated HbA1C during illness. A Smartphone based system is developed that allows patients to monitor BG concentrations and report symptoms of illness and illness. Data gathered by patients through use of this device, will be used to test the hypothesis that changes in daily BG measurements can be used to predict illness in T1DM patients, before symptoms onset. A successful prediction model will enable patients to get early indication of upcoming illness, before they are bedridden. Patients can thus actively take precautions to avoid or shorten illness episodes or make these less severe and/or have healthy BG concentrations during illness. This project is breaking new grounds by detecting illness before the onset of symptoms and illness, and an illness prediction model using patient observable parameters will be an important advance in the field of disease surveillance and prediction.