1023-P: Interaction of COVID-19 and Preexisting Diabetes on Hospital Presentation of Diabetic Ketoacidosis

Background: Larger studies are needed to evaluate whether COVID-19 increases the risk of DKA presentation. Methods: We examined individual and joint associations of COVID-19 and preexisting diabetes (DM) with DKA presentation in a retrospective cohort of patients who presented to Montefiore Medical Center in the Bronx, NY from March to June 2020. We defined DKA as acidosis (pH 250 mg/dl. We defined COVID-19 status by PCR test result and preexisting DM by HbA1c ≥6.5%, ICD-10 code, or outpatient DM medications. Given likely bias in hospital presentation during the pandemic, a historical control from March to June 2019 was included in the COVID negative group. Data were analyzed using multivariate regression. Results: We included 46,479 patients: mean age 43 yrs, 56% male, 57% with DM of which 94% type 2 DM. COVID-19 alone was not associated with higher odds of DKA presentation, though as expected preexisting DM was. Adjusted regression analyses for interaction showed 64% higher odds of DKA for patients with both COVID-19 and preexisting DM than the additive effect of each condition alone (Table 1). Conclusions: COVID-19 has a multiplicative effect on the risk of hospital presentation of DKA in people with preexisting DM. Mechanistic studies examining DKA in COVID-19 in people with preexisting DM may help identify preventive targets. Disclosure J. Mathew: None. C. Schechter: None. W. Southern: None. J. P. Crandall: None. Y. Tomer: None. S. Agarwal: None.