Comparing the resolution of metabolic acidosis in severe diabetic ketoacidosis in patients receiving Plasmalyte-148 compared to isotonic sodium chloride 0.9%

Diabetic ketoacidosis (DKA) represents a major complication of insulin-dependent diabetes, where significant osmotic diuresis is driven by renal excretion of glucose. As such, these patients have the potential to lose large fluid volumes, causing metabolic acidosis, electrolyte abnormalities, and considerable dehydration. DKA-associated volume depletion requires aggressive rehydration, and isotonic sodium chloride commonly represents the crystalloid of choice in guidance. However, infusion of large saline volumes potentiates hyperchloraemic metabolic acidosis, thus risking delayed treatment resolution.

Considering the above, Ramanan et al.[1] compared Plasmalyte-148 (PL) against conventional sodium chloride 0.9% (SC) in patients with severe DKA. Criteria for severe DKA were defined as arterial serum pH<7.25 (or serum bicarbonate <15mmol/L), serum glucose >14mmol/L, and admission necessitated to ITU.

The primary aim of the study was:

• To compare the speed of resolution of severe DKA by measuring normalisation of base excess (base excess > -3mEq/L at 48hrs post ICU admission).

Secondary aims considered:

• Monitoring blood ketone concentration and whether PL would mitigate resolution, owing to the acetate found in PL.
• Evaluating median length of ITU and hospital stay.
• Contrasting adverse effects of fluid therapy (hypoglycaemia and electrolyte imbalances).

Of the 93 patients enrolled in the study, sourced from 7 different ITU settings, 90 were included in the final analysis and randomised across the two arms (PL vs SC). Overall, the baseline demographics and biochemical profiles were similar in the two groups. However, the PL group had a greater percentage of type 1 diabetes mellitus (83% vs 67%) and different distribution of base excess, despite similar means. In addition, compliance measured the median proportion of assigned-to-total fluid received; this was lower in the PL group (66% vs 100%).

96% and 86% achieved the primary outcome in the PL and SC groups, respectively, with base excess being significantly higher in the PL group (+3.5 mEq/L, 95% CI 0.4–6.7, p=0.026). Blood ketone concentration remained similar in each group throughout. Median lengths of stay in ITU and hospital were shorter in the PL group. However, these did not appear statistically significant. Hypophosphataemia (<0.7mmol/L) occurred less commonly in the PL group (69% vs 81%), as well as hypoglycaemic events (<3.8mmol/L; 19% vs 26%) and moderate-severe hypokalaemia (<3.0 mmol/L; 10% vs 17%).

STUDY STRENGTHS & LIMITATIONS

Strengths

• The study design is well-described, transparent, and appropriate for the trial. The crossover cluster design allowed the study to be conducted across multiple ITU settings (rural and city), generating a greater sample size.
• The results presented are easy to comprehend, suggestive, and relate clearly to the outcomes specified.
• The authors do well in recognising a potential complication of PL, monitoring ketone-body generation appropriately, and thus improving confidence in using PL for future studies.

Limitations

• The authors decided to measure DKA resolution using base excess, a parameter not used in isolation as a marker of resolution; this makes real-life implantation of main outcomes more difficult.
• The study lacks significant power, owing to the small sample size and its pre-set design as a phase II clinical trial.
• Open-label study risking bias.
• Volume and rate of fluids determined by treating clinician – risk of bias.
• Only patients admitted to ITU were included in the study, limiting the degree of data extrapolation to hospital-wide patients.
• There was disparate compliance between the two arms, potentially limiting the effects that PL would have had in isolation. Similarly, variability in results may have occurred from differences in baseline diabetes status.
• The authors did not discuss potential complications of resolving DKA too rapidly, fluid shifts, and associated cerebral oedema. These are important complications to remember for future interventions.

TAKE-HOME MESSAGES

The use of Plasmalyte-148 may accelerate the resolution of base excess, compared to saline 0.9%, in patients with severe DKA admitted to ITU. Patients receiving Plasmalye-148 may also have shorter lengths of ITU and hospital stay and fewer adverse biochemical outcomes. This data, however, is not conclusive and requires further investigation.

This article review was prepared and submitted by Dr Steven Schofield (FY2) and Dr Andrew Martin (Consultant in ICM & Anaesthesia – Manchester Royal Infirmary, Manchester (UK).