Poster Presentation Australian Diabetes Society and the Australian Diabetes Educators Association Annual Scientific Meeting 2017

A retrospective evaluation of prevalence and management of hypokalaemia in a cohort of patients with diabetic ketoacidosis (#273)

Kay Hau Choy 1 , Florence Gunawan 1 , Mark A Kotowicz 1 2 3
  1. Department of Endocrinology and Diabetes, Barwon Health, Geelong, Victoria, Australia
  2. Deakin University, Geelong, Victoria, Australia
  3. Melbourne Medical School - Western Campus, The University of Melbourne, Victoria, Australia

Objectives:

Hypokalaemia is a potentially serious complication of diabetic ketoacidosis (DKA). Based on National Safety and Quality Health Standards (NSQHS), infusions of potassium are restricted to ≤20mmol/L on wards at University Hospital Geelong. We assessed the prevalence of hypokalaemia in adult patients with DKA at our tertiary hospital.

Methods:

A retrospective medical record review of adult patients, age of ≥18yr, with a DKA diagnosis admitted between January 1, 2016 to December 31, 2016 was conducted. Diagnosis of DKA was based on, pH<7.35, blood glucose>11.0mmol/L and capillary ketones>1.0. Hypokalaemia was defined as serum potassium <3.5mmol/L and mild, moderate, severe and critical hypokalaemia were defined as 3.0-3.4mmol/L, 2.5-2.9mmol/L, 2.0-2.4mmol/L and <2.0mmol/L, respectively. Resolution of ketoacidosis was defined as pH>7.3 with capillary ketone<1.0.

Results:

Of 149 adult patients identified, 68 (mean age 42.6±19.8yrs) met all criteria (88 % type 1 diabetes, 12% type 2 diabetes). Hypokalaemia at presentation occurred in 5.9% (4/68) (mean±SD serum potassium 4.7±0.98mmol/L). Hypokalaemia developed in 54.4% (37/68) in the first 48 hours. Hypokalaemia was mild in 39.7%, moderate in 10.3%, severe in 1.5% and critical in 2.9%. The mean±SD potassium replacement administered within the first 48 hours was 149±148.3mmol, with 115±82.2mmol of potassium replacement in the first 24 hours. Intensive care unit (ICU) admission occurred in 36.8% (25/68) with median length of stay in ICU being 2 days. Resolution of DKA within 24 hours occurred in 95.6% and median length of hospital stay was 3 days. No cardiac arrhythmia was observed and there was no in-hospital mortality.

Conclusion:

The prevalence of hypokalaemia on presentation was relatively high and increased 10-fold in the first 48 hours of treatment. Hypokalaemia was mostly mild to moderate but 4.4% experienced severe to critical hypokalaemia, suggesting insufficient potassium replacement and that NSQHS guidelines may need to be modified to permit safe treatment of DKA.