A Retrospective Analysis of the Clinical Characteristics and Treatment Patterns Among Individuals with Diabetes Receiving Insulin Therapy in Japan: The Insulin JP2DB Study

INTRODUCTION: This retrospective database study investigated trends in insulin and use of concomitant noninsulin glucose-lowering medication (NIGLM) among Japanese individuals with diabetes. METHODS: The study comprised two analyses: (1) a serial cross-sectional analysis of patterns in insulin treatment by year (database: Real World Data); and (2) a longitudinal retrospective cohort analysis that examined insulin treatment and individuals' characteristics (database: DeSC). Individuals initiating insulin in an inpatient or outpatient setting were followed up for 9 months (type 2 diabetes [T2D]) or 21 months (type 1 diabetes [T1D]) to evaluate treatment changes over time. RESULTS: The serial cross-sectional analysis included 4953 individuals (T2D, n = 4693; T1D, n = 260). The proportion of participants with T2D receiving concomitant NIGLMs increased from 31% in 2002 to 61% in 2021; from 2014 onwards, more than 30% of insulin-treated and basal-insulin-treated individuals treated with concomitant NIGLMs received dipeptidyl peptidase-4 inhibitors. Since 2018, use of concomitant NIGLMs in T1D has increased. The longitudinal retrospective cohort analysis included 27,492 individuals (T2D, n = 27,031; T1D, n = 461). Among participants with T2D who initiated insulin in an inpatient setting, 70.8% received bolus insulin at initiation, with this proportion declining to 17.3% after 9 months; proportions of participants receiving basal insulin and of those receiving basal-bolus insulin increased over the same period (6.3-31.1% and 17.0-23.4%, respectively). The majority of participants with T2D who initiated insulin in an outpatient setting received basal insulin at initiation (hospital, 53.9%; clinic, 58.9%). Among participants with T1D who initiated insulin in an inpatient setting, 57.9% received bolus insulin, and basal-bolus insulin was the predominant regimen after 1 month (85.0%); in outpatient settings, basal-bolus insulin was the predominant regimen throughout the study. CONCLUSION: In Japan, the most prominent insulin regimen at initiation varied across settings in T2D but not in T1D; use of concomitant NIGLMs increased over time in both.