Incident Diabetes Diagnosis among the Uninsured during COVID-19 Hospitalization.

Introduction: Undiagnosed diabetes is associated with lack of insurance, which decreases access to preventive care. During the COVID-19 pandemic, uninsured patients previously unknown to health systems were hospitalized.

Methods: This is a cross-sectional analysis of electronic health record (EHR) data from patients with diabetes hospitalized with COVID-19 in a safety-net health system from June 2020 to December 2021, examining the association between payor status and incident diagnosis of diabetes. Incident diagnosis of diabetes was defined by excluding a prior known diagnosis of diabetes based on diagnosis codes, medications, and hemoglobin A1c from the past 5 years. Regression analysis evaluated the association between payor status and incident diagnosis of diabetes. Data were analyzed in 2023.

Results: Among 872 patients with diabetes hospitalized with COVID-19, 24.0% were uninsured, 34.6% received county-funded charity coverage (CFCC), 17.1% received Medicaid, and 24.3% received Medicare. The rate of incident diagnosis of diabetes in the total sample was 20.3%; incident diagnosis of diabetes was more common among the uninsured (30.1%) than CFCC (18.2%) and Medicare patients (11.3%). After adjusting for age, gender, race/ethnicity, and BMI, uninsured patients had higher odds of incident diagnosis of diabetes (aOR 2.64, 95% CI [1.41-4.92], p=0.002) than Medicare patients. Odds of incident diagnosis of diabetes were similar for CFCC and Medicare patients.

Conclusion: Uninsured patients had higher odds of incident diagnosis of diabetes during COVID-19 hospitalization that may have gone undetected without hospitalization. These findings reflect decreased access to preventive care and missed opportunities to screen for diabetes among the uninsured.