Abstract: Bioinformatic Evaluation of the Child Opportunity Index 2.0 and Other Sociodemographic Factors as Predictors of Diabetes Clinic Appointment Completion in Youth with Type 1 Diabetes
Page Clements1, Kelsey Panfil1, Hamp Henning1, Brent Lockee1, David D. Williams1, Erin M. Tallon1, Anna R. Kahkoska2, Angelica Cristello Sarteau2, Susana R. Patton3, Mark A. Clements1
1Children’s Mercy Kansas City (Kansas City, MO), 2University of North Carolina at Chapel Hill (Chapel Hill, NC), 3Nemour's Children's Health, (Jacksonville, FL)
Background and Aims: Canceled or missed diabetes clinic appointments impact the health of youth with type 1 diabetes (T1D), though little is known about factors that underlie this population’s risk for missed appointments. We examined relationships between appointment completion and age, race, ethnicity, insurance status, and the Child Opportunity Index 2.0 (COI; a street address-derived proxy for neighborhood opportunity and access to quality socioeconomic and educational resources).
Methods: We analyzed electronic health records (EHR) of youth (T1D duration >6 months) receiving care from a pediatric diabetes clinic network in the Midwest USA between January 2016 and August 2023. Youths’ addresses were geocoded to census tracts which correspond to COI scores standardized at the national level (values ranging 0-100). We used mixed-effects logistic regression to predict clinic appointment completion with COI, race, ethnicity, insurance, and age at appointment as fixed effects and youth as a random effect.
Results: We evaluated data for 26,644 appointments (501 canceled/158 missed) for 3,411 youth (52.4% male; 80.6% White; 64.4% commercial insurance; median age 14.4 years [5.7 IQR] with median COI score 67 [45 IQR]). Factors that independently increased the probability of appointment completion included White race (vs. Black or African American and Other races; Log Odds = 0.2, p<.05) and higher COI (Log Odds = 0.01, p<.0001).
Conclusions: In a bioinformatic analysis, higher-quality neighborhood conditions were associated with higher likelihood of completing clinic appointments. Future studies should evaluate COI in advanced, clinically deployable models to predict missed appointments, along with interventions to reduce them.