Abstract
Background: Modeling techniques in the field of pediatrics present unique challenges beyond traditional model limitations, and sometimes difficulties in faithfully simulating the condition's evolution over time. Objective: This study aimed to identify whether economic modeling approaches in diabetes in pediatric patients align with the recommendations of clinical practice guidelines and the latest scientific evidence. Methods: A literature review was performed in March 2023 to identify modeling-based economic evaluations in diabetes in pediatric patients. Data were extracted and synthesized from eligible studies. Clinical practice guidelines for diabetes were gathered to compare their alignment with modeling strategies. Two endocrinology specialists provided insights on the latest findings in diabetes that are not yet included in the guidelines. A multidisciplinary group of experts agreed on the relevant themes to conduct the comparative analysis: parameter informing on glycemic control, diabetic ketoacidosis/hypoglycemia, C-peptide as prognostic biomarker, metabolic memory, age at diagnosis, socioeconomic status, pediatric-specific sources of risk equations, and pediatric-specific sources of utilities/disutilities. Results: Nineteen modeling-based studies (7 de novo, 12 predesigned models) and 34 guidelines were selected. Hemoglobin A1c was the main parameter to model the glycemic control; however, guidelines recommend the usage of complementary measures (eg, time in range) which are not included in economic models. Eight models included diabetic ketoacidosis (42.1%), 16 included hypoglycemia (84.2%), 2 included C-peptide (1 of those as prognostic factor) (10.5%) and 1 included legacy effect (5.3%). Neither guidelines nor models included recent findings, such as age at diagnosis or socioeconomic status, as prognostic factors. The lack of pediatric-specific sources for risk equations and utility/disutility values were additional limitations. Discussion: Economic models designed for assessing interventions in diabetes in pediatric patients should be based on pediatric-specific data and include novel adjuvant glucose-monitoring metrics and latest evidence on prognostic factors (C-peptide, legacy effect, age at diagnosis, socioeconomic status) to provide a more faithful reflection of the disease. Conclusions: Economic models represent useful tools to inform decision making. However, further research assessing the gaps is needed to enhance evidence-based health economic modeling that best represents reality.