Effect of Low Bolus Frequency on Automated Insulin Delivery System Performance in Youth With Type 1 Diabetes.

Abstract

To assess differences in time in range (TIR) between automated insulin delivery (AID) systems in youth with low bolus frequency. Youth using Omnipod 5 (OP5) or Tandem Control-IQ (CIQ) averaging ≤3.0 boluses/day over 90 days were included. Propensity score matching accounted for between-group differences. A multiple linear regression model assessed for differences in TIR by system. A total of 202 youth had low bolus frequency (n = 75 of 290 using CIQ, n = 127 of 535 using OP5; P = 0.5). Propensity score matching paired 98 youth for analyses and eliminated between-group differences, excluding time in AID, which remained higher for CIQ users (80.0% vs. 64.0%; P = 0.01). There was no between-group difference in user-initiated boluses (2.2 boluses/day). Adjusted TIR was 8.0% higher in CIQ users (54.9% vs. 46.9%; P = 0.002). Among AID users receiving ≤3.0 boluses/day, CIQ use was associated with greater TIR, likely due to increased time in AID and AID-initiated boluses.