# Metabolic Complications During the Initial Management of Diabetic Ketoacidosis According to Diabetes Type in a Single Tertiary Centre in Malaysia

## TL;DR

This study compares how type 1 and type 2 diabetes patients experience complications during treatment for diabetic ketoacidosis, finding differences in hypoglycemia and hospital outcomes.

## Contribution

The study identifies diabetes-specific metabolic risks and clinical outcomes during DKA management, suggesting the need for tailored treatment approaches.

## Key findings

- T1DM patients had higher hypoglycemia rates compared to T2DM patients.
- T2DM patients experienced longer hospital stays and more infections.
- Tailoring DKA management to diabetes type may reduce complications.

## Abstract

Introduction: Diabetic ketoacidosis (DKA) is traditionally associated with type 1 diabetes mellitus (T1DM) but is increasingly observed in type 2 diabetes mellitus (T2DM). Despite notable pathophysiological differences between these conditions, current DKA management follows standardized protocols regardless of diabetes classification. This “one-size-fits-all” approach raises concerns about potential treatment-related metabolic complications that are specific to distinct diabetes phenotypes.

Objectives: To compare the rates of metabolic complications - specifically hypoglycemia and hypokalemia - during the initial management of DKA between T1DM and T2DM patients and identify associated factors. Secondary outcomes included evaluating the length of hospital stay (LOS), complication rates, and in-hospital mortality, and identifying predictors of hypoglycemia and hypokalemia during DKA treatment.

Methods: A retrospective cohort study of DKA admissions at Penang General Hospital, Malaysia, was conducted between 2019 and 2023. Data on DKA admissions, metabolic complications, and hospitalization outcomes were extracted and analysed.

Results: We reviewed the case records of 215 DKA admissions: 100 (46.5%) T1DM, 115 (53.5%) T2DM. The prevalence of hypoglycemia was significantly higher in patients with T1DM compared to T2DM (33/100, 33.0% vs 18/115, 15.7%, p=0.003). Conversely, the prevalence of hypokalemia did not differ significantly between the groups (55/100, 55.0% vs 62/115, 53.9%, p=0.873). Multivariate analysis identified T1DM and low body weight as independent predictors of hypoglycemia. Independent predictors for hypokalemia included low pH, low initial serum potassium, shorter diabetes duration, and absence of nephropathy. Patients with T1DM had a lower observed mortality rate (0/100, 0.0% vs 4/115, 3.5%) and fewer 30-day readmissions (4/100, 4.0% vs 10/115, 8.7%) compared to T1DM, but not statistically significant. T1DM patients required more intensive care, whereas T2DM patients experienced significantly longer hospital stays and higher rates of nosocomial infection.

Conclusion: Significant differences exist in both metabolic safety profiles and clinical trajectories when managing T1DM and T2DM patients with DKA. While T1DM patients are prone to hypoglycemia, T2DM patients carry a higher burden of comorbidities leading to prolonged hospitalization and infection risks. Tailoring DKA management strategies specifically to diabetes type may reduce treatment-related complications and improve overall clinical outcomes.