Discussion
In recent years, there has been a renewed focus on the type of rehydration fluid in DKA that might influence the time to acidosis resolution and mitigate complications of hyperchloremia, hypokalemia and new-onset/progressive AKI. There is emerging data to suggest the deleterious effects of NS on acid-base balance as well as the causation of AKI in critically ill children and adults.18–22 Despite this, determining the ideal type of fluid still remains an unmet need and an area of active research in DKA management.14 The recent clinical practice consensus guidelines recommend NS for volume expansion; however, the choice of crystalloid for subsequent deficit correction lacks strong consensus and can be 0.45% saline, NS or balanced crystalloids like RL or plasmalyte.23
Problems related to DKA management in RLS are myriad. Poor compliance with insulin, delayed health-seeking, misdiagnosis and late referral increase the risk for severe DKA and complicate AKI. The time to resolution depends on the severity of DKA at presentation.24 The median time to DKA resolution reported from low- and middle-income countries (LMICs), where the proportion of children presenting with severe DKA is high, ranges between 14.5 and 17 hours.7 25 In contrast, developed countries, where the proportion of children presenting with severe DKA is lower, reported shorter resolution times ranging between 8 and 14 hours.8 Given this background, large volumes of chloride-liberal fluids in RLS may delay DKA resolution by causing hyperchloremia and AKI. It was, therefore, imperative to evaluate whether balanced fluids, by virtue of their lower chloride content and ability to regenerate
bicarbonate ions, could mitigate these problems. In our trial, the time to DKA resolution was lower in the RL as compared with the NS arm. Although initial single-center studies comparing balanced fluids with NS failed to show significant differences in children and adults,7 17 recent multicenter, pragmatic, cluster cross-over randomized controlled trial, SCOPE-DKA (Sodium Chloride or Plasmalyte-148 Evaluation in Diabetic Ketoacidosis) in adults have demonstrated faster resolution of DKA in the balanced crystalloid group.5 Further evidence supporting the use of balanced fluids came from subgroup analysis of two large pragmatic companion cluster randomized controlled trials (SMART {Isotonic Solutions and Major Adverse Renal Events Trial} and SALT-ED {Saline against Lactated Ringer’s or Plasma-Lyte in the Emergency Department}), which included adults with DKA presenting to the emergency department. RL constituted 97% of the balanced fluids used. The median (IQR) time to DKA resolution in the balanced fluid group was significantly shorter, 13 (9.5–18.8) hours, compared with the NS group, which was 16.9 (11.5–34.9) hours.26 A recently published trial from India that compared RL with NS in children with DKA demonstrated shorter PICU stay and less hyperchloremia in the RL group.9 Yung et al showed that in the subgroup of children with severe DKA, the use of Hartmann’s fluid as compared with NS led to earlier DKA resolution.8 These findings compare favorably with the results of our current study.
(Sodium Chloride or Plasmalyte-148 Evaluation in Diabetic Ketoacidosis) in adults have demonstrated faster resolution of DKA in the balanced crystalloid group.5 Further evidence supporting the use of balanced fluids came from subgroup analysis of two large pragmatic companion cluster randomized controlled trials (SMART {Isotonic Solutions and Major Adverse Renal Events Trial} and SALT-ED {Saline against Lactated Ringer’s or Plasma-Lyte in the Emergency Department}), which included adults with DKA presenting to the emergency department. RL constituted 97% of the balanced fluids used. The median (IQR) time to DKA resolution in the balanced fluid group was significantly shorter, 13 (9.5–18.8) hours, compared with the NS group, which was 16.9 (11.5–34.9) hours.26 A recently published trial from India that compared RL with NS in children with DKA demonstrated shorter PICU stay and less hyperchloremia in the RL group.9 Yung et al showed that in the subgroup of children with severe DKA, the use of Hartmann’s fluid as compared with NS led to earlier DKA resolution.8 These findings compare favorably with the results of our current study.
Furthermore, two recently published meta-analyses, predominantly including adults with DKA, favored the use of balanced fluids over NS.27 28 A recent systematic review and network meta-analysis that compared the effectiveness of fluid schemes in children with DKA also supported the use of RL over NS.29 As against this good-quality evidence in adults, evidence in children with respect to balanced fluids in DKA is limited.
We found significant hyperchloremia at 4 and 8 hours of therapy in the NS group, similar to the observations of several previous studies.5 6 17 30 The rise in bicarbonate from baseline to 12 hours was significantly higher in the RL compared with the NS group, again an observation reported by other trials.6 30 The hyperchloremic effect of NS in the initial hours and the rise of bicarbonate in the RL group form a strong physiological basis for early DKA resolution in the RL group.
The incidence of AKI in DKA, at the time of presentation, is reported to be as high as 50%, most of which resolves with adequate hydration.31 We observed similar high rates of AKI (49%) in our cohort, all responding to hydration, indicating a prerenal cause for its occurrence. The prevalence of AKI at admission and new-onset/progressive AKI was similar in both groups. Given the background of high prerenal AKI in children with DKA in RLS, whether a second hit due to hyperchloremia could cause new/progressive AKI/delayed recovery of kidney function and future risk of CKD is yet to be conclusively established. In the meantime, using chloride-restrictive balanced fluid like RL until more evidence is generated, especially in high-risk situations, may be a viable option. Hypokalemia and hypoglycemia tend to be more frequent and severe in undernourished DKA in RLS.32 More than 50% of children in this study group developed at least one episode of hypokalemia during treatment of DKA, although the occurrence was similar between the two groups.
Leave a Reply