Linkage to Treatment Among Admitted Individuals With Diabetes Identified Through Routine Emergency Department Screening

Diabetes remains an epidemic, disproportionately affecting populations with limited access to care, who also may rely on the emergency department (ED) for their medical needs. Early diagnosis and treatment can prevent severe complications such as kidney failure and limb loss. While it is known that ED populations have high rates of undiagnosed diabetes, outcomes of routine ED screening on initiation of diabetes care are unknown. This study aims to characterize linkage to care (LTC) outcomes among admitted patients with new diagnoses of diabetes.

A large urban ED in Chicago implemented routine, automated hemoglobin A1C (HbA1C) screening for all individuals undergoing a complete blood count who: 1) had no known diabetes, 2) had no screening within 3 years, and 3) had body mass index ≥ 25 kg/m2 or age ≥ 35. This is a retrospective chart review of all patients with HbA1C 6.5% (indicative of diabetes) ordered in the ED between May 8, 2024, and May 7, 2025, who were subsequently admitted to the hospital. Discharged patients were excluded as they had a separate LTC pathway. Individuals were considered linked if they were scheduled a follow-up appointment for diabetes management prior to discharge. Logistic regression was used to evaluate predictors of LTC.

During the study period, 1,861 individuals met study criteria, of whom 15% had no previous diabetes diagnosis. Of all patients, 972 (52%) were successfully linked to care. Those with new diagnoses were linked at lower rates (47%) than those with known diabetes who were out of care (56%). Patients with a new diagnosis who were not linked had significantly lower HbA1C values (mean 7.5%) compared to those who were linked (mean 9.4%) and those with known diabetes who were linked (mean 9.1%). Of patients linked to care, 49% did not have an existing primary care provider. Overall, men had 24% lower odds of LTC compared to women (p = 0.004), and patients admitted to Cardiology, Surgery, or Neurology had 41-66% lower odds of LTC (p < 0.001) compared to Internal Medicine.

ED-based diabetes screening identified large numbers of new and poorly controlled diabetes, suggesting that this could be a critical public health intervention that would successfully address gaps in access to care among vulnerable populations. Further research is needed to improve LTC, including examination of both patient and healthcare system factors.