The Type of Patient Training Does Not Impact Outcomes in the First 90 Days of Automated Insulin Delivery Use.

Background: Youth starting Omnipod 5 (OP5) can onboard with a diabetes educator or self-start with support from online, industry-provided educational modules. We compared glycemic control and pump interaction by training type among youth initiating OP5. Methods: This retrospective review included 297 youth with type 1 diabetes (T1D) aged <22 years initiating OP5. We analyzed baseline continuous glucose monitor (CGM) data and pump and CGM data from the first 90 days of OP5 use. Multilevel mixed-effects regression assessed for changes in time in range (TIR) from baseline to 90 days by training type. Results: Of youth initiating OP5, 42.4% trained with a diabetes educator and 57.6% self-started. At baseline, self-starters had a longer T1D duration (5.0 (2.6,7.9) vs. 2.5 (1.3, 5.5) years, P = 0.001), more time <54 mg/dL (0.3% (0.1,1) vs. 0.15% (0,1), P = 0.01), and a higher coefficient of variation (40.2% (37, 44.4) vs. 38.7% (34.4, 42.4), P = 0.004). After 90 days of OP5 use, groups did not differ in time in automated mode or boluses per day. In a longitudinal model, after adjusting for baseline TIR and T1D duration, 90-day TIR was 10.5%-points higher (CI: 9.2-11.8, P < 0.0001), positively associated with baseline TIR (β = 0.82, CI: 0.78-0.85, P < 0.0001), and 1.1%-points greater among self-starters (CI: 0.06-2.2; P = 0.04). Conclusions: After 90 days of OP5 use, glycemic control and pump interactions were minimally different between youth who self-started and those who trained with a diabetes educator. For youth at a tertiary care center previously using an Omnipod system, online educational modules offered by industry provide sufficient training for use.