Carbohydrate count aided by a simulation of postprandial glucose dynamics generated by a mathematical model in people with type 1 diabetes mellitus

Abstract

ABSTRACT Diabetes mellitus (DM) is one of the most common non-communicable chronic diseases. Its prevalence is increasing, and the increase is expected to be more marked in poor and developing countries. For people with type 1 diabetes mellitus (DM1), insulin is the only medication approved for treatment, and for people with type 2 diabetes mellitus (DM2). However, there are new medications; in countries like Colombia, the use of insulin is quite frequent. A treatment based on reducing blood glucose concentration reduces the risk of chronic complications, and carbohydrate counting is the most widely recommended dose calculation technique. Carbohydrate counting is an iterative process for people with DM accompanied by high uncertainty, especially initially. Carbohydrate counting is refined with experience and frequent interventions by doctors and patient support staff. This work aimed to probe whether carbohydrate counting aided by a simulation was better than the traditional technique to estimate insulin dose in people with DM1. To do this, a phenomenological mathematical model of the glucose dynamics in the gastrointestinal system was built to simulate the changes experienced by an individual's blood glucose after a meal. Then, a mobile and simple-to-use solution was developed so that a person could simulate a meal and evaluate the impact of insulin dose estimated by carbohydrate counting. We hypothesized that frequent use of a solution like the one described should result in better glucose control. This hypothesis was tested in a randomized controlled clinical trial.