Towards patient-specific carbohydrate counting accuracy: an in silico study

Abstract

Type 1 diabetes mellitus patients on intensive insulin therapy use advanced carbohydrate counting to proper dose prandial insulin. Therefore, the patient’s ability to accurately estimate the meal’s carbohydrate content is paramount. However, despite its significance, several studies show that the patient’s ability to estimates the meal’s carbohydrate content is far from ideal and identify the need for continuous education on carbohydrate counting. In this context, the authors have proposed in previous works an analytic method to determine the maximum error to the carbohydrate counting regarding each patient’s insulin-to-carb ratio and the insulin sensitivity factor. This maximum can be of great significance to design patient-specific educational programs and to define learning outcomes according to the specific characteristics of each patient. This work presents a methodology and conditions to assess the previously proposed method, using the FDA-approved University of Virginia(UVA)/Padova Type 1 Diabetes Simulator.