Incidence and predictors of diabetic ketoacidosis among adult type 1 and type 2 diabetes mellitus patients at public hospitals in Harari Region, eastern Ethiopia: A retrospective cohort study

Background Diabetic ketoacidosis (DKA) is the most prevalent and serious acute complication of diabetes mellitus. Over the past decade, the global incidence of DKA hospitalizations has risen, with recent studies reporting a 55% increase. Therefore, this study aimed to assess the incidence and identify predictors of DKA among adult patients with diabetes in eastern Ethiopia. Methods An institution-based retrospective cohort study was conducted in public hospitals in the Harari region of Ethiopia from January 1, 2019, to December 31, 2024, among 455 adults with diabetes mellitus. Data collection was performed using the Kobo toolbox, and analysis was carried out using STATA software version 17. The Cox proportional hazards regression model was applied to identify predictors of DKA. Adjusted hazard ratios (AHR) with 95% confidence intervals (CI) and corresponding p-values were computed. Results Out of the 446 patients included in the study, 110 (24.66%), 95%CI (20.87%-28.88%) developed diabetic ketoacidosis. The incidence rate of DKA was 1.1 cases per 100 person-months (95% CI: 0.9–1.3), with rates of 2.7 per 100 person-months for T1DM and 0.7 per 100 person-months for T2DM. Medication non-adherence (AHR: 2.27, 95% CI: 1.46, 3.54), poor glycemic control (AHR: 2.79, 95% CI: 1.72, 4.54), acute febrile illness (AHR: 2.15, 95% CI: 1.51, 3.07), urinary tract infection (AHR: 3.04, 95% CI: 1.99, 4.64) and overweight (AHR: 2.23, 95% CI: 1.45, 3.42) were predictors significantly associated with DKA. Conclusion The study revealed that diabetic ketoacidosis occurred in one out of four diabetic patients. Factors such as medication non-adherence, poor glycemic control, overweight, acute febrile illness, and urinary tract infections significantly increased the risk of DKA. Therefore, targeted follow-up care is essential for diabetic patients with these identified predictors to reduce the incidence of DKA.