Acute kidney injury and length of ICU stay in diabetic ketoacidosis: a retrospective cohort study.

AIMS: Diabetic ketoacidosis (DKA) is a major acute complication of diabetes mellitus and a leading cause of diabetes-related morbidity. Acute kidney injury (AKI) further compounds this burden, yet data on AKI and prolonged intensive care unit (ICU) stay in DKA from the Middle East remain scarce. We aimed to evaluate AKI prevalence and identify predictors of AKI and prolonged ICU stay among adults admitted to the ICU with DKA at a Jordanian tertiary hospital. METHODS: This retrospective cohort study included 183 adults admitted to the ICU with DKA (2012-2024) at King Abdullah University Hospital, Jordan. AKI was defined by Kidney Disease: Improving Global Outcomes (KDIGO) creatinine criteria; prolonged ICU stay was defined as a total ICU length of stay of ≥3 days. Multivariable logistic regression identified independent predictors. RESULTS: AKI occurred in 15.9% of patients and 28.7% had prolonged ICU stay. Ninety-day mortality was 4.4%. Elevated admission glucose independently predicted AKI [adjusted odds ratio (aOR) 1.05 per mmol/L; 95% confidence interval (CI) 1.00-1.10; p = 0.034]. Predictors of prolonged ICU stay were AKI (aOR 4.88; 95% CI 1.73-13.77; p = 0.003), infection (aOR 4.00; 95% CI 1.83-8.76; p < 0.001), elevated sodium (aOR 1.07; 95% CI 1.00-1.14; p = 0.039), and lower potassium (aOR 0.52, 95% CI 0.28-0.97, p = 0.040). Higher sodium levels, AKI incidence, and infection were robust across the ≥2-day threshold, infection was robust also at the ≥4-day threshold, while potassium was significant only at the ≥3-day definition. CONCLUSIONS: Among ICU-admitted patients with DKA, AKI affected one in six and was a strong independent predictor of extended hospitalization. Admission hyperglycemia, infection, and hypernatremia are accessible markers that may support early risk stratification, though these exploratory associations warrant prospective validation.