IDF21-0589 Initial learnings from an eHealth platform for identifying individuals at high risk of diabetes in Mexico

Abstract

Background: Type 2 diabetes mellitus (DM2) is a leading cause of morbidity and mortality in Mexico. Since the early onset of DM2 is associated with a worse prognosis, the widespread implementation of evaluation risk tests could facilitate its identification among subjects at high risk, in order to promote strategies that reduce the risk and burden of the disease. Aim: To report the initial findings of the open access eHealth platform implementation, based on a chat-box, for the identification of risk factors for diabetes and subjects at high risk of the disease. Method: To identify risk factors for diabetes, an open access eHealth platform (GYANT®) was promoted among the general population in Mexico, through social networks of non-governmental organizations (Mexican Diabetes Federation). This platform is a chat-bot that uses artificial intelligence to guide the conversation with spontaneous users, who provide the information requested. Afterwards, the platform provides the user with a result on their risk profile for diabetes and suggests practical medical care recommendations in a friendly way. Data obtained from January to October 2020 were analyzed to evaluate the performance of this initiative and identify subjects at high risk. The diabetes risk test (DRT) and the metabolic syndrome cohort office-based prediction model (MSCPM) for incidental diabetes were used to define the presence of diabetes risk. Today, the MSCPM is the only predictive model for diabetes in the Mexican population and was developed in middle-aged adults. Results: Data from 23,662 participants were analyzed. The mean age of the participants was 21 ± 9 years, most of them male (51%), with secondary or higher education (79%). The risk factors in the participants were: family history of diabetes (99.1%), overweight (26.1%), hypertension (20%) and obesity (19.8%). Moderate to high risk of diabetes was found in 12%, according to the diabetes risk test, and in 75.4% for the MSCPM (moderate risk [69.1%], moderate-high risk [5.4%] and high risk [0.9%]). A higher BMI was associated with self-reported hypertension (13%), hypercholesterolemia (12.6%), smoking (12%), and poor exercise performance (14.6%). Discussion: Participants in the eHealth platform reported a high prevalence of high-risk factors for diabetes. Early age and a family history of diabetes may have led to participation, which indicates about health concerns that occur in this young adult population. This is the first report of its kind on health initiatives to identify the risk of diabetes in Mexico. We acknowledge that the use of self-reported data is a limitation of this study, as it may be less reliable due to possible lack of precision in responses, as well as possible intentional and unintentional reporting bias. However, we consider that this type of eHealth tools could help to identify subjects with high-risk profiles, so its massive implementation should be considered in health policies, including binding actions to the strategies of diagnosis and/or prevention programs.