Endocrine Abstracts

Introduction

Once thought to typify type 1 diabetes mellitus (T1DM), diabetic ketoacidosis (DKA) in people with type 2 diabetes mellitus (T2DM) is also now increasingly recognized. We explored the clinical and biochemical differences in DKA in the two types of diabetes.

Methods

This retrospective cohort study included all the DKA episodes from April 2014 to September 2020 at a large tertiary care centre in West Midlands, United Kingdom. People admitted with DKA were classified as having T1DM or T2DM based on previously established diagnosis, treatment history, autoantibody status, and/or phenotypic features. The two groups were compared for differences in precipitating factors, biochemical profiles at presentation, management, and outcomes of the admission with DKA.

Results

A total of 786 DKA episodes were identified. Eighteen were excluded due to unclear underlying type of diabetes. 75.9% (n = 583/768) had T1DM, and 24.1% (n = 185/768) had T2DM. The most common precipitating causes in both groups were intercurrent illness (total: n = 272/768, 35.42%; T1DM: n = 198/583, 33.96%; T2DM: n = 74/185, 40.00%) and suboptimal compliance to treatment (total: n = 206/768, 26.82%; T1DM: n = 176/583, 30.19%; T2DM: n = 30/185, 16.22%). There was no clear precipitant in 16.15% (n = 124/768) (T1DM: n = 95/583, 16.30%; T2DM: n = 29/185, 15.68%). There was no significant difference in biochemical profiles on admission [median(Q1-Q3) pH (T1DM: 7.22(7.09–7.29); T2DM: 7.24(7.11–7.30); P = 0.3266), bicarbonate (T1DM: 11.90(7.13–16.78); T2DM: 13.20(7.75–17.80); P = 0.2192), glucose (T1DM: 28.00(20.45–34.80); T2DM: 26.55(16.21–35.09); P = 0.4496), lactate (T1DM: 2.6(1.8–4.3); T2DM: 2.6(2.04–4.21); P = 0.6532), serum osmolality (T1DM: 310.07(300.61–320.90); T2DM: 312.13(300.11–328.79); P = 0.2787)] between the two groups, except for urea, which was higher in T2DM [T1DM 7.1(5.1–10.6); T2DM 8.9(6.4–16.8); P < 0.0001]. There was no difference in appropriateness of fixed rate intravenous insulin infusion administered [T1DM: 100.00%(93.53–104.48); T2DM: 99.48%(90.00–102.27); P = 0.0688], proportion of hourly glucose measurements [T1DM: 99.49%(61.71–131.95); T2DM: 92.78% (64.33–120.22); P = 0.2143], proportion of hourly ketone measurements [T1DM: 46.39% (24.30–71.43); T2DM: 44.43% (20.90–70.74); P = 0.6485], and fluids administered [T1DM: 80.00% (58.33–100.00) vs 80.00% (50.00–112.50); P = 0.4978]. People with T1DM had more episodes of hypoglycaemia [T1DM 0(0–1); T2DM 0(0–0); P = 0.0056] during DKA management. Although there was no difference in the total DKA duration between the two groups [T1DM: 13.92(9.11–21.92); T2DM: 13.90(7.73–21.12); P = 0.4638], people with T2DM had significantly longer hospital stay [T1DM: 2.95(1.68–6.05); T2DM: 11.02(4.99–23.11)]; P < 0.0001].

Conclusion

Both cohorts had similar DKA severity, management, complications, and duration. People with T2DM required longer hospital stay, suggesting a more complex need for care, and increased economic burden. Increased awareness of DKA in people with T2DM is needed for improved prediction and prevention.