Prandial Insulin Versus Glucagon-Like Peptide-1 Added to Basal Insulin: Comparative Effectiveness in the Community Practice Setting

Abstract

Background: Real-world data on emerging combination approaches for type 2 diabetes mellitus (T2DM) management are limited. The objective of the current study was to document the characteristics and clinical outcomes of patients with T2DM initiating prandial insulin or a glucagon-like peptide-1 (GLP-1) receptor agonist while on basal insulin. Methods: This was a retrospective analysis of an electronic medical records database of patients with T2DM managed in a community practice setting in the United States. The main outcome measures were glycated hemoglobin (HbA(1c)), body weight, hypoglycemia, and health care resource utilization at baseline and at 6-month and 1-year follow-up. Results: A total of 33 810 patients were included in the study: 31 848 on prandial insulin and 1962 on a GLP-1 receptor agonist. At baseline there were significant differences in mean age (60 vs 56 years), mean Charlson Comorbidity Index score (1.1 vs 0.7), mean HbA(1c) (8.8% vs 8.4%), and mean body weight (99 vs 112 kg) between the prandial insulin and GLP-1 receptor agonist groups, respectively (P < 0.001 for each). After matching for baseline differences, significant and similar changes from baseline were observed between the prandial insulin and the GLP-1 receptor agonist groups during follow-up at the 6 months/1 year post-index date for HbA(1c) (-0.45/-0.60% vs -0.44/-0.58%, respectively; P = 0.907/0.723 between groups). Body weight changes between the groups were significantly different at 6 months/1 year (+1.7/-1.7 vs -0.9/-3.7 kg; P < 0.001).Hypoglycemia incidence and health care resource utilization were significantly greater in the prandial insulin versus GLP-1 receptor agonist group. Conclusions: The results of this real-world analysis of patients with T2DM adding a GLP-1 receptor agonist or prandial insulin to basal insulin suggest an association between adding a GLP-1 receptor agonist with similar glycemic control, greater reduction in body weight, lower hypoglycemia incidence, and lower health care utilization compared with adding prandial insulin