Key points are not available for this paper at this time.
Background & Objective: The risk of new onset diabetes following infection with COVID-19 was contrasted in two distinct cohorts. Methods: Using Merative MarketScan Commercial Claims database (MarketScan) and electronic health records from University of Kentucky HealthCare (UKHC), we created three cohorts in each database: 1) a COVID-19 infected cohort, 2) a contemporaneous cohort of individuals without COVID-19 infection, and 3) a historic pre-COVID-19 cohort (Table 1). Diabetes incidence rates and hazard ratios were contrasted between the COVID-19 cohort and two control cohorts (no-COVID & pre-COVID). Results: Compared to the commercial database, diabetes prevalence was higher in UKHC database, reflecting an objectively sicker population (higher comorbidity index and increased hospitalizations). In both databases, diabetes incidence rates and hazard ratios were significantly higher in the COVID cohort (Table 1), compared to either pre-COVID or non-COVID cohorts. Conclusions: Both databases demonstrate that COVID-19 was associated with 2-3-fold increased rate of incident diabetes. However, as compared to an academic medical center with a sicker patient population, relatively healthier patients in the commercial data set had lower incidence of COVID-19 associated diabetes. Disclosure M.E. Lacy: None. A.P. Smith: None. L.R. Hammerslag: None. K. Heier: None. A.B. Fields: None. J. Talbert: None. J. Fowlkes: None. S.J. Fisher: None. P.A. Kern: None. Funding National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health (R01DK124626)
LACY et al. (Fri,) studied this question.