Characteristics and outcomes of DKA in people taking SGLT2 inhibitors: DEKODE study

Nitish Mathur; Angelica Sharma; Nevil Philip; Lakshmi Rengarajan; Lucy Bomphrey; Anu Abraham; Sanjay Saraf; Ketan Dhatariya; Punith Kempegowda; Group Dekode Working

doi:10.1530/endoabs.99.OC12.4

OC12.4

Prev Next

Section Contents Cite

Endocrine Abstracts (2024) 99 OC12.4 | DOI: 10.1530/endoabs.99.OC12.4

ECE2024 Oral Communications Oral Communications 12: Diabetes, Obesity, Metabolism and Nutrition | Part II (6 abstracts)

Characteristics and outcomes of DKA in people taking SGLT2 inhibitors: DEKODE study

Nitish Mathur ¹ , Angelica Sharma ² , Nevil Philip ³ , Lakshmi Rengarajan ^3,4 , Lucy Bomphrey ⁵ , Anu Abraham ³ , Sanjay Saraf ⁶ , Ketan Dhatariya ^2,7 , Punith Kempegowda ^3,4 & Dekode Working Group ⁴

Views

Author affiliations

¹The Royal Wolverhampton NHS Trust, Diabetes and Endocrinology, Wolverhampton, United Kingdom; ²Norfolk and Norwich University Hospitals NHS Foundation Trust, Norwich, United Kingdom; ³The University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom; ⁴Institute of Applied Health Research, Clinical Trials Unit College of Medical and Dental Sciences, University of Birmingham, Birmingham, United Kingdom; ⁵Walsall Manor Hospital, Walsall, United Kingdom; ⁶Good Hope Hospital, Birmingham, United Kingdom; ⁷University of East Anglia, Norwich, United Kingdom

Background: Although SGLT2 inhibitors (SGLT2-i) have revolutionised the management of diabetes, they are associated with an increased risk of developing diabetic ketoacidosis (DKA). However, there remains a paucity of evidence about the risk factors, management, and outcomes of DKA episodes associated with SGLT2-i therapy. Improved understanding of these factors in order to optimise education, minimise risk, and improve outcomes of DKA episodes in individuals receiving SGLT2-i therapy is important.

Objective: To study the characteristics, causes and complications associated with the management of DKA in patients who were on SGLT2-i therapy vs those who were not.

Methods: We retrospectively analysed all DKA episodes across six hospitals participating in the DEKODE (Digital Evaluation of Ketosis and Other Diabetes-related Emergencies) initiative between January 2020-December 2022. Parameters including demographics, DKA precipitants, laboratory values, duration of DKA (hours), length of hospital stay (days), and mortality were collated and compared among individuals who developed a DKA episode on SGLT2-i vs those who did not. Statistical analyses were performed using SPSS 29.0.

Results: Of 1398 DKA episodes, 6% (n=84) developed DKA whilst on SGLT2-i therapy. Of these, 19% (n=16/84) were directly attributable to SGLT2-i therapy. SGLT2-i users were older (58 years vs 41 years, P<.001) and more commonly of Afro-Caribbean ethnicity (15.5% vs 6.2%). At presentation, SGLT2-i users had a lower pH (7.17 vs 7.20, P=.024), bicarbonate (10.7 vs 12.0 mmol/l, P=.039), glucose (18.4 vs 27.6 mmol/l, P<.001) and ketone (5.5 vs 6.0 mmol/l, P=.018) concentrations. COVID infections (13.1% vs 3.9%, P=.00007) accounted for a greater proportion of DKA episodes amongst SGLT2-i users. SGLT2-i users developed more episodes of hypokalaemia during DKA management (46.4% vs 33.9%, P=.024). However, the development of hyperglycaemia and hyperkalaemia during DKA management were comparable in patients with or without SGLT2-i therapy. DKA episodes whilst on SGLT2-i therapy had a longer duration (19.6 hrs vs 14.9 hrs, P=.001), prolonged hospital admission (5.4 days vs 3.4 days, P<.001) and higher mortality (7.1% vs 2.7%, P=.048).

Conclusion: DKA episodes in individuals receiving SGLT2-i therapy presented with lower glucose and ketone concentrations but more severe acidosis. These episodes were associated with more adverse outcomes compared to DKA in individuals not treated with SGLT2-i therapy.