Endocrine Abstracts

Background/aims: Although patients with diabetic ketoacidosis (DKA) are expected to have total body potassium depletion, measured levels may be normal or elevated due to extracellular shifts of potassiumsecondary to acidosis. We aimed to examine the prevalence of hypokalemia in patients with DKA at presentation and during the treatment.

Methods: This is a retrospective cross-sectional descriptive study concerning all patients hospitalized in the Endocrinology Department for DKA between October 2021 and January 2022. Initial kalemia levels (k+) on presentation and those after treatment were determinated on AU 680^® Beckman Coulter. Hypokalemia was defined using the criteria set forth by the American Diabetes Association (ADA) as measured serum potassium less than 3.3 mmol/l. The data were analyzed using SPSS26.

Results: A total of 30 patients were hospitalized for DKA. The mean age of the group was 26.73±10.41 years. The mean potassium level was 3.73 mmol/l (range 2.3 to 5.3; S.D. ± 0.72). Among patients, 26.66%had hypokalaemia (k+< 3.3 mmol/l), 16.66% had k+ between 3.3 and 3.5 mmol/l. In 26.66% of cases the k+ was between 3.5 and 4 mmol/l and 16.66% of patients had k+ between 4 and 4.5 mmol/l. In 10% of patients the k+ was between 4.5 and 5 mmol/l. Only one patient was presented with hyperkalemia. The mean age of the group who presented hypokalemia was 27 125 years ± 8.62 and 50% of them were Type 1 diabetes. None of the patients had neuromuscular or electrical signs of dyskalemia. During treatment, the mean k+ was 3.43 mmol/l (range 2.5 to 4.1; S.D. ± 0.4). The prevalence of hypokalemia was 33.33%. Among the patients who had hypokaliemia at presentation, 75% of them developed an hypokaliemia during treatment. And among the patients who had normal values of k+, 18.18% of them developed an hypokaliemia during treatment. The mean of duration of insulin therapy were 38.5 (± 24.21) hours and 30.69 (±18.77) hours (P=0.039) respectively for the group who presented hypokalemia and for the group with normal k+ during the treatment. Also the group who presented hypokalemia during the treatment received higher total insulin dose infusion (85.09±47.30 units vs 52.66 units ± 29.80);(P=0.039).

Conclusion: Hypokalemia was observed in 26.66% of patients with DKA. Further research is needed to better determine the risks and benefits of administering insulin before obtaining serum potassium values. The strategy of lowering insulin infusion rate in patients with significant hypokalemia during DKA treatment should require further evaluation.