Background: T1D autoantibodies can be detected months to years before disease onset, but few people undergo screening. We explored attitudes toward autoantibody screening in people with T1D and caregivers and relatives of people with T1D. Methods: Participants were recruited from the T1D Exchange Registry or referred by an Online Registry participant. Guided by the Health Belief Model, we conducted 4 focus groups with 26 participants (11 adults with T1D with a child without T1D; 8 caregivers to a child with T1D and a child without T1D; 7 biological siblings to a person with T1D). Transcripts were analyzed using NVivo. Results: Responses are summarized in Table 1. Most participants held positive attitudes toward screening. The most common perceived benefit was to obtain knowledge; the anticipated emotional burden of a positive screen was the most frequently reported barrier. Anxiety and relief were emotions associated with a positive and negative screen, respectively. Participants desired information about antibody screening and its interpretation and accuracy. A healthcare providers’ recommendation might prompt the decision to screen. Conclusions: Participants expressed positive attitudes and perceived benefits of autoantibody screening but also reported barriers to screening, particularly the emotional burden of a positive result. Disclosure M.Peter: None. K.S.M.Chapman: None. J.L.Dunne: Employee; Janssen Research & Development, LLC. C.S.Kelly: None. W.Wolf: None.

Research on diabetes distress in continuous glucose monitor (CGM) users with T1D is divided. Some studies suggest CGM users experience less distress compared to non-users, whereas other studies find no differences in distress. While research shows higher distress is related to higher HbA1c, more work is needed to understand how distress is related to CGM metrics. Thus, we recruited adult CGM users with T1D (n = 199) from the T1D Exchange Registry to complete an online survey about diabetes characteristics and a screener of diabetes distress (DDS-2). CGM metrics (defined in Table 1) were computed from CGM data within 12 months prior to survey completion. Participants were grouped into lower (DDS-2 < 3, n = 120) or higher (DDS-2 ≥ 3, n = 79) diabetes distress. We used Welch’s t-tests to compare mean differences in blood glucose (BG) experiences and CGM metrics between participants with lower and higher diabetes distress. As shown in Table 1, participants with higher diabetes distress reported higher HbA1c, and - via CGM metrics - experienced more time in higher BG ranges and less time within standard target ranges, had higher blood glucose variability, and had higher GMI compared to those with lower distress. Associations between distress and CGM metrics replicated in a MANCOVA after adjusting for duration of T1D. Our results suggest, in CGM users, higher perceived diabetes distress is associated with more time in hyperglycemia. Disclosure C. S. Kelly: None. H. Nguyen: None. K. S. M. Chapman: None. E. M. Cornelius: None. M. Peter: None. W. Wolf: None.

Use of automated insulin delivery (AID) in people with type 1 diabetes (T1D) has increased in recent years. We analyzed data from the T1D Exchange Registry - an online longitudinal study following adults and children with T1D - to describe characteristics of AID use and its association with self-reported HbA1c, occurrence of diabetic ketoacidosis (DKA) symptoms, and occurrence of severe hypoglycemia events (SHE) . Of 12,065 participants (69.5% female, 87.5% non-Hispanic White, mean age 37.9 yrs, mean T1D duration 19.9 yrs) , 26.4% reported using AID, 43.6% using insulin pump without AID, and 30.0% using multiple daily injections (MDI) . Private insurance was reported in 77.4% of AID users, 74.9% pump without AID, and 66.3% MDI. Among the 3,185 AID users, 50.6% used predictive low glucose suspend (PLGS) features, including Medtronic 640G and Tandem Basal-IQ; 49.4% used hybrid closed-loop (HCL) , including DIY looping, Medtronic Auto Mode, and Tandem Control-IQ. Average HbA1c was lower in AID users than those using MDI and pump without AID; Incidence of SHE was lower among HCL users compared with PLGS users and AID users compared with MDI users (Table 1) . These cross-sectional real-world data support findings of previous clinical studies showing significant improvements in HbA1c with AID use, with HCL use providing additional protection against SHE. Disclosure J.Liu: None. W.Wolf: None. K.Miller: Research Support; Dexcom, Inc., Tandem Diabetes Care, Inc. C.Kelly: None. K.S.M.Chapman: None. M.Peter: None. D.Finan: None. H.Nguyen: None. K.Laferriere: None. C.Leon: None.

Diabetes outcomes are affected by socioeconomic factors including health insurance. While some research has examined the impact of health insurance in people with type 1 diabetes (T1D) (e.g., disruptions in health insurance on health outcomes) , it is less clear whether health insurance type (e.g., private versus government-sponsored) shares a unique association with diabetes outcomes. Thus, we examined relationships between demographics, socioeconomic status, and health insurance type with HbA1c in adults in the T1D Exchange Registry - an online longitudinal study of people living with T1D. Participants (N = 7725) were 42.2 years old (SD = 14.6) on average and 75.4% reported their gender as female. Most participants (92.2%) reported their race as White and 5.1% identified as Hispanic. Many participants used insulin pumps (71.3%) and continuous glucose monitors (CGM, 83.0%) . Self-reported HbA1c was 7.3% (SD = 1.58) on average. Health insurance type was categorized into private (79.0%) , Medicaid (9.6%) , Medicare (9.5%) , or no health insurance (1.9%) . Health insurance type, demographic factors (i.e., age, gender, race, ethnicity) , socioeconomic factors (i.e., income, education level) , and diabetes technology use (i.e., insulin pump, CGM) were entered simultaneously into a linear regression with HbA1c as the outcome variable. Private insurance was used as the reference group. Statistical significance was set at p < .05. We found lower HbA1c was associated with identifying race as White (Β = -0.21) , having higher income (Β = -0.10) , having higher education (Β = -0.17) , using an insulin pump (Β = -0.46) and using CGM (Β = -0.72) . Higher HbA1c was associated with identifying as female (Β = 0.14) and having either Medicaid (Β = 0.23) or no health insurance (Β = 0.30) . These results suggest, above and beyond other socioeconomic and demographic factors, health insurance type may contribute to small but meaningful differences in HbA1c. Disclosure C.Kelly: None. H.Nguyen: None. J.Liu: None. K.S.M.Chapman: None. M.Peter: None. C.Leon: None. K.Laferriere: None. J.Ravelson: None. W.Wolf: None.