Examining the reach of a diabetes screening program in an urban emergency department

Abstract

Background

Screening for diabetes in non-traditional settings like emergency departments (ED) can enhance early detection among patients at higher risk for diabetes. This study aims to assess the reach of an ED-based screening program by examining the characteristics of patients screen-detected for diabetes or prediabetes.

Study design

Retrospective cross-sectional study.

Methods

Sociodemographic characteristics (age, gender, race and ethnicity, insurance, zip code) of patients who were screened for diabetes using hemoglobin A1c test (A1c) were examined. The distribution of prediabetes and diabetes within each race/ethnicity, age, and gender groups were determined. ArcGIS Pro 2.9.0 was used to geocode patient zip codes, to generate heat maps of high occurrences of prediabetes and diabetes.

Results

Of the 5997 individuals screened in the ED, 49 % were non-Hispanic Black, 27 % Hispanic, 15 % non-Hispanic White, 5 % non-Hispanic Asian, and 4 % non-Hispanic Other/unknown. Almost half (47 %, n = 2808) had elevated A1c levels indicative of prediabetes (n = 2070; A1c: 5.7–6.4 %) or diabetes (n = 738; A1c: ≥6.5 %). Non-Hispanic Black females had a higher prevalence of both prediabetes (54 %) and diabetes (55 %) diagnoses as compared to other race/ethnic or gender categories; whereas non-Hispanic Asians had a lower prevalence of both prediabetes and diabetes except for those ≥65 years or older. Furthermore, most patients screened for prediabetes and diabetes reside in resource poor neighborhoods on the west and south sides of Chicago.

Conclusion

The burden of prediabetes and diabetes were greater among non-Hispanic Black females, with a high prevalence of prediabetes observed among younger individuals, particularly those residing in resource poor neighborhoods in the west and south sides of Chicago. More investment in resources for diabetes prevention and management for these groups may be warranted.