A Predictive Model for an Assistant System to identify Diabetes Mellitus

Given the high incidence in Mexico of Diabetes Mellitus (DM), it seems impossible for all patients to be monitored by a specialist. Therefore, preventive action and primary care contact services are an urgent need for early detection of this disease. This paper aims at proposing a model to calculate the risk factor from the information provided by the participant without requiring clinical studies. Firstly, the principal causes of DM are identified via an initial state-of-art analysis about DM pre-screening, screening, diagnosis, and treatment. Then, a digital survey was applied to a sample of fifty participants, all of whom were asked about their knowledge of DM’s causes. Afterward, an in-depth qualitative analysis of the answers was made. Results showed that only 36% of the interviewed subjects identified the main symptoms of DM, but 66% would be willing to consult with a doctor if they could identify the symptoms. The proposed digital system measures the perception of the user and provides quick information about the disease, considering DM Type 1 statistics. The system is based on conditional probability and clinic guides for prevention by identifying risk factors and pondering. Further extensive use of this assistance could raise awareness about the risk factor and provided a percentage of the probability of having DM Type 1, and the result could help detect DM in the early stages.