E-Poster Presentation ESA-SRB-ANZBMS 2021

Hypophosphataemia in association with the treatment of diabetic ketoacidosis: predictive significance of severity of acidosis at presentation (#320)

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

Objectives: Hypophosphataemia is commonly associated with the treatment of diabetic ketoacidosis (DKA). However, literature on the dynamics of serum phosphate and determinants of hypophosphataemia in DKA remains scarce. We evaluated the change in serum phosphate during DKA treatment and assessed whether acidosis severity on admission was related to the degree of hypophosphataemia.

Methods: A retrospective review of adult patients (aged ≥18 years) with DKA (pH<7.35, blood glucose >11mmol/L and capillary ketones >1.0mmol/L) admitted between September 2020—July 2021. Those without a serum phosphate (Pi) at presentation or serial Pi during DKA treatment were excluded.

Results: Of 75 DKA episodes, 36 in 28 patients, mean±standard deviation (SD) age 45.4±20.2yrs, 75% type 1 and 25% type 2 diabetes, met inclusion criteria. At presentation, 58.3% were hyperphosphataemic (Pi>1.50mmol/L). Initial Pi (mean±SD 1.68±0.61mmol/L) correlated with initial serum glucose (r =0.522, P=0.001), but not serum creatinine (r =0.185, P=0.281). Pi decreased during treatment in all cases (mean±SD Pi reduction 1.19±0.67mmol/L, mean±SD nadir Pi 0.49±0.25mmol/L), with a hypophosphataemic (<0.75mmol/L) nadir Pi in 83.3% and severe hypophosphataemia (<0.32mmol/L) in 16.7%. Initial serum bicarbonate correlated with nadir Pi (r =0.582, P<0.001). Using linear regression, every 1.0mmol/L decrease in serum bicarbonate was associated with an average reduction of 0.05mmol/L in Pi. Mean±SD initial bicarbonate differed between those with a nadir Pi <0.5mmol/L and ≥0.5mmol/L (8.1±4.6mmol/L vs 13.1±6.3mmol/L, P=0.022), whereas those with severe hypophosphataemia had a mean±SD bicarbonate of 7.8±4.4mmol/L vs 10.7±6.0mmol/L (P=0.270) compared with those with nadir Pi ≥0.32mmol/L. No adverse effects of hypophosphataemia on morbidity or mortality were noted.

Conclusion: There is a significant relationship between the degree of metabolic acidosis at presentation and the extent of hypophosphataemia during DKA therapy. Clinicians should be vigilant of the risk of severe hypophosphataemia during DKA treatment in patients with profound acidosis, particularly in those with a serum bicarbonate of <10mmol/L.