What is the clinical evidence from this study?

Study design: Cross-Sectional. Population: Type 1 and Type 2 Diabetes. Intervention: EHR-based surveillance. Primary outcome: Diabetes prevalence by type (T1D and T2D).

What does this research mean for the field?

Electronic health record-based estimates of childhood diabetes prevalence by type and demographic subgroup are consistent with traditional population-based surveillance, showing T1D is most prevalent in non-Hispanic White children and T2D in non-Hispanic Black and Hispanic children. Novelty: ClaimNovelty.CONFIRMATORY. Consensus alignment: ConsensusAlignment.SUPPORTS_CONSENSUS.

What question did this study set out to answer?

To estimate the prevalence of diabetes types in US children using electronic health records in 2018 and 2022.

June 7, 2026

2344-P: Electronic Health Record (EHR)-Based Estimates of Diabetes Prevalence by Type in Children: The Diabetes in Children, Adolescents, and Young Adults Network (DICAYA), 2018 and 2022

Key Result

In 2018, EHR-based prevalence of type 1 diabetes was 3.8 per 1000 children in health systems and 2.3 in geographic centers, while type 2 diabetes prevalence was 1.5 and 0.4 per 1000, respectively.

Key Points

To estimate the prevalence of diabetes types in US children using electronic health records in 2018 and 2022.
Used data from 5 centers in the DiCAYA Network spanning two US states.
Identified children under 18 years with type 1 or type 2 diabetes using validated computable phenotypes.
Calculated 3-year period prevalence for health system centers and 1-year point prevalence for geographic-based centers.
In 2018, type 1 diabetes prevalence was 3.8 per 1000 in health system centers and 2.3 in geographic centers.
Type 2 diabetes prevalence was noted as 1.5 in health system and 0.4 in geographic centers.
Prevalence was higher among non-Hispanic White for type 1 and among non-Hispanic Black and Hispanic children for type 2.

Study Design

Type

Cross-Sectional

Multicenter

Yes

Structured PICO

Population

Children <18 years old in the US evaluated for type 1 or type 2 diabetes prevalence using EHR data from 5 centers in 2018 and 2022.

Outcome

Diabetes prevalence by type (T1D or T2D) and demographic subgroup in index years 2018 and 2022

EHR networks offer a scalable approach for timely surveillance of childhood-onset diabetes, yielding prevalence estimates consistent with recent population-based surveillance.

Abstract

Introduction and Objective: Childhood-onset diabetes is associated with costly care and reduced life expectancy. In US surveillance data, prevalence rose by ~50% from 2001 to 2017. EHRs may offer a scalable approach for timely surveillance. Our objective was to use EHR networks to estimate diabetes prevalence by type and demographic subgroup in US children in index years 2018 and 2022. Methods: We used data from 5 centers (Table 1) in the DiCAYA Network: 3 health system centers (HS) and 2 geographic-based centers (GBC) in two US states. Validated computable phenotypes identified children 18y with type 1 (T1D) or type 2 (T2D) diabetes in the index year or up to several preceding years. We used bias correction to adjust HS denominator estimates. We calculated a 3-year period prevalence for HS and a 1-year point prevalence for GBC. Results: In 2018, T1D prevalence was 3.8 per 1000 children 18y (95% CI: 3.7, 3.9) in the HS and 2.3 (2.2, 2.3) in the GBC. T2D prevalence was 1.5 (1.4, 1.5) in the HS and 0.4 (0.4, 0.4) in the GBC. Table 1 shows results for 2018 and 2022 by type and subgroup. T1D was most prevalent among non-Hispanic (NH) White and T2D among NH Black and Hispanic children. Prevalence increased with age. T2D was more prevalent in girls than in boys. Conclusion: In this large EHR-based study of children, diabetes prevalence findings by type and demographic subgroup were consistent with recent population-based surveillance. Disclosure L. Zhang: None. A.G. Hirsch: None. T. Crume: None. A.D. Liese: None. R. Anthopolos: None. H. Shao: None. Y. Guo: None. B.E. Dixon: None. D. Dabelea: None. S. Conderino: None. E.A. Mendonca: None. M. Mefford: Research Support; Current; Merck U18DP006512; U18DP006509; U18DP006500; U18DP006513; U18DP006506; U18DP006693; U18DP006694; U18DP006517; U18DP006518; and U18DP006633.

Bookmark