1961-LB: Evaluation of Two Behavioral Economic Interventions for Youth with Type 1 Diabetes—A Randomized Controlled Trial

Introduction 7.2%, and/or predicted to have a 90-day rise in A1C of 0.3% via machine learning-based model. Subjects were randomized into 3 groups: COIN2DOSE (C2D) - uses positive financial incentives to promote mealtime insulin use, with weekly bonuses for sustained meal dosing; LOAN2DOSE (L2D) - uses a loss aversion approach to promote mealtime insulin use; or control. The potential financial compensation was equal for C2D and L2D over the 12-week intervention period. Results: To date, 32 youth have enrolled and 24 have completed the study mean age= 14.8±1.4 years, mean T1D duration= 7.1± 3.2 years, mean A1C=8.9±1.6%, 58% female, 9 C2D, 6 L2D, 9 control. At baseline, there were no differences in A1C, time in range (TIR), or mealtime insulin use (BOLUS) by group. Mean BOLUS score showed a non-statistically significant increase from 0.93±0.93 to 1.3± 0.92 in C2D baseline to post-treatment, which was sustained at 6-months post-baseline (p=0.64). Mean BOLUS score increased from 1.1±0.1to 1.4±1.0 in L2D baseline to post-treatment but decreased to 0.86±0.15 at 6-months post-baseline (p=0.82). The control group BOLUS score increased from 1.3±1.0 to 1.4±0.9 at post-treatment and 1.5±0.78 at 6-months post-baseline (p=0.86). Across groups, there were no changes in TIR and A1C at post-treatment or 6-months post-baseline. Conclusion: Preliminary results suggest that either positive financial incentives or loss aversion may be associated with short-term improvement in mealtime insulin use, but not A1C or TIR. We continue the trial to enable more testing of group differences and long-term impact. Disclosure S. Tsai: Stock/Shareholder; Dexcom, Inc. R. McDonough: Speaker's Bureau; Provention Bio, Inc. S.R. Patton: None. B. Spartz: None. D.D. Williams: None. B. Lockee: None. M.A. Clements: Research Support; Abbott. Consultant; Glooko, Inc. Research Support; Dexcom, Inc. Funding Helmsley Charitable Trust