P2.D070: Hyperkalemia During Diabetic Ketoacidosis and the Association With Acute Kidney Injury

Background: Potassium abnormalities are common during diabetic ketoacidosis (DKA). While hypokalemia has been attributed to renal losses, hyperkalemia has largely been attributed to insulin deficiency. However, acute kidney injury (AKI), which is an increasingly recognized as a complication of DKA, may also contribute to potassium retention. Our primary aim was to determine whether there is an association between hyperkalemia and AKI during episodes of DKA. Our secondary aim was to determine factors associated with hyperkalemia during DKA.

Methods: We conducted an observational retrospective cohort study among patients presenting to a pediatric emergency department with a diagnosis of DKA between January 1st, 2012 – September 30th, 2023. Hyperkalemia was defined as the first non-hemolyzed serum potassium ≥5.5 mEq/L. AKI was defined by the Kidney Disease Improving Global Outcomes creatinine criteria using the Schwartz estimation equation to calculate an expected baseline creatinine. DKA was defined as having all of the following: 1) pH < 7.3 or bicarbonate < 15 mEq/L, 2) positive urine or serum ketones, and 3) glucose >200 mg/dL. We excluded encounters with missing laboratory values, hemolyzed potassium, or transfer from an outside hospital. The Mann Whitney U test with Yate's correction and Chi-square were used for univariate analysis. Binary logistic regression was used to test the association between hyperkalemia and AKI, adjusting for covariates, including age, sex, new-onset vs. known type 1 diabetes (T1D), pH, glucose, blood urea nitrogen, and hemoglobin A1c.

Results: We identified 1,197 eligible encounters with DKA, of which 114 (9.5%) had hyperkalemia and 230 (19.2%) had AKI. Of encounters with hyperkalemia, 114 (100%) received insulin, 2 (1.8%) received intravenous calcium, 1 (0.9%) received albuterol, and 21 (18.4%) had an electrocardiogram. The most common electrocardiogram finding was sinus tachycardia (n=9) followed by peaked T-waves (n=7). One encounter had a wide-complex tachycardia. The proportion of encounters with AKI was significantly higher among encounters with hyperkalemia (n =53, 47.3%) compared to without hyperkalemia (n=177, 16.5%) (p-value < 0.001; Table 1). AKI was independently associated with hyperkalemia after adjusting for covariates (aOR 1.88, 95% CI 1.12, 3.16; Table 2). Known compared to new-onset T1D (aOR 3.14, 95% CI 1.77, 5.58), and severe compared to mild acidosis (aOR 5.25, 95% CI 2.71, 10.15) were associated with greater odds of hyperkalemia. A one standard deviation increase in serum glucose was associated with greater odds hyperkalemia (aOR 1.56, 95% CI 1.26, 1.93), and a one standard deviation increase in hemoglobin A1c was associated with lower odds of hyperkalemia (aOR 0.71, 9%% CI 0.56, 0.90).

Conclusion: In this single-center study, DKA-related AKI was independently associated with hyperkalemia. Known T1D, more severe acidosis, and higher glucose were associated with greater odds of hyperkalemia, whereas higher hemoglobin A1c was associated with lower odds of hyperkalemia.