Pediatric DKA Management: 5 Critical Practice Considerations
After the PECARN FLUID trial, what the evidence shows about rehydration rate, saline tonicity, and the risk of cerebral injury in children.
One-Minute Takeaway
- Contemporary pediatric DKA management no longer treats intravenous fluids as the primary cause of cerebral injury, and the strongest evidence on this question is a large randomized controlled trial.
- The PECARN FLUID trial randomized 1,389 diabetic ketoacidosis episodes in children; neither a faster nor slower rehydration rate, and neither saline concentration, changed neurologic outcomes.
- Decline in Glasgow Coma Scale score did not differ by rehydration rate (RR 0.76, 95% CI 0.44–1.33), and clinically apparent brain injury occurred in roughly 0.9% of episodes.
- Observational studies tie cerebral injury to illness severity — low PCO₂ (RR 3.4, 95% CI 1.9–6.3) and high blood urea — rather than to fluid infusion rate (RR 1.1, 95% CI 0.4–3.0).
- Because cerebral injury can begin before treatment starts, current guidelines prioritize restoring perfusion over withholding fluids.
Why Pediatric DKA Management Matters
Diabetic ketoacidosis is among the most common life-threatening emergencies in children with type 1 diabetes, and it accounts for most diabetes-related deaths in childhood. Of every complication clinicians weigh, cerebral injury is the one they fear most.
Clinically overt cerebral injury complicates roughly 0.5% to 0.9% of episodes, yet its consequences are severe. Early case series reported mortality between 40% and 90%. In more recent cohorts, about one in five affected children died, and a similar proportion survived with lasting neurologic deficits.
The scale explains the caution. Ketoacidosis occurs in roughly a quarter to two-fifths of children at the onset of type 1 diabetes. Cerebral injury, in turn, has accounted for about half to two-thirds of diabetes-related deaths in childhood.
For decades, the dominant concern in pediatric DKA management was that intravenous fluids themselves triggered this injury. That belief pushed protocols toward cautious, slow rehydration. The practical result was a standing tension: give enough fluid to restore perfusion, or give less for fear of harm.
Both errors carry cost. Too little fluid risks prolonged hypoperfusion, acute kidney injury, and slower resolution; too much has been blamed for the edema clinicians most want to avoid. Pediatric DKA management has spent years trying to locate the safe middle of that range.
The Clinical Problem in Pediatric DKA Management
The core question in pediatric DKA management is direct. Does the way fluids are given cause cerebral injury, or merely coincide with it?
Early physiological reasoning held that rapid fluid administration lowered plasma osmolality quickly, drawing water into brain cells and producing swelling. On that basis, many protocols advocated slow rehydration with isotonic fluid and a large assumed deficit.
A competing model gained ground. Cerebral injury may reflect hypoperfusion and ischemia–reperfusion, amplified by the inflammatory state of ketoacidosis, with edema a consequence rather than the initiating cause.
Pathophysiological reviews now describe a two-hit sequence. The first hit is vasogenic edema from disrupted blood–brain barrier permeability during the ketoacidotic, inflammatory state; the second is cytotoxic swelling as osmolality shifts during treatment. In this framework, fluids act at most as a secondary modifier of an injury that ketoacidosis has already set in motion.
Supporting evidence spans clinical and laboratory work. Imaging studies show reduced cerebral blood flow during untreated ketoacidosis followed by hyperemia during treatment, and animal models reproduce swelling driven by ischemia and inflammation rather than by fluid osmolality alone. This shifts the emphasis of pediatric DKA management from what is infused toward the underlying disease process.
These observations reframed the problem. If injury can precede fluids and tracks illness severity, then fluid strategy may matter far less than long assumed — a hypothesis that demanded a randomized trial to settle.
Available Approaches
Three practical choices define fluid strategy in DKA: how fast to rehydrate, what tonicity to use, and how large a deficit to assume. Each has been framed, historically, around the same worry — cerebral injury — which makes the fluid decision a hinge of pediatric DKA management. The table summarizes the evidence that has tested these choices against that risk.
| Study / Source | Design | N | Population / Setting | Focus | Key result on fluids and cerebral injury |
|---|---|---|---|---|---|
| Kuppermann 2018 (PECARN FLUID) | Randomized controlled trial, 2×2 factorial | 1,389 episodes (1,255 children) | 13 US pediatric emergency departments; age ≤18 y | Rate (fast vs slow) and saline (0.9% vs 0.45%) | No difference in GCS decline — fast vs slow RR 0.76 (0.44–1.33); 0.45% vs 0.9% RR 0.80 (0.46–1.40); brain injury 0.9% overall with no between-group difference |
| Glaser 2001 | Multicenter case–control | 61 cases; 355 controls | 10 US centers, 1982–1997 | Biochemical and treatment risk factors | Fluid infusion rate not associated (RR 1.1, 0.4–3.0); low PCO₂ (RR 3.4, 1.9–6.3), high BUN (RR 1.7, 1.2–2.5), and bicarbonate (RR 4.2, 1.5–12.1) associated |
| Edge 2006 | National case–control (UK) | 43 cases; 169 controls (from 2,940 episodes) | UK population surveillance | Baseline and treatment risk factors | Severe acidosis strongly associated (OR 0.02, 0.002–0.15, least vs most acidotic); early insulin (OR 4.7, 1.5–13.9) and higher early fluid volume also associated |
| Lawrence 2005 | Population-based nested case–control | 13 cases; 26 controls | Canada, 1999–2001 | Incidence and risk factors | Incidence 0.51%; associated with low bicarbonate, high urea, and high glucose; no association with any treatment factor |
All rows were verified against primary sources this session; every DOI appears in the references and was confirmed to resolve.
Conservative (slow) rehydration
The traditional approach assumed a large fluid deficit — often 10% of body weight — replaced evenly over about 48 hours, with no additional bolus beyond initial resuscitation. Its rationale was osmotic caution: correct slowly to avoid rapid shifts.
The strongest test of this approach is the slow arm of the PECARN FLUID trial. It produced neurologic outcomes no better than faster rehydration, which undercuts the assumption that slower is inherently safer.
The approach persisted less because trials supported it than because the alternative felt risky. Absent randomized data, restriction became the cautious default, and protocols encoded it for a generation of clinicians.
More liberal (rapid) rehydration
The faster PECARN regimen assumed the same deficit but replaced half within the first 12 hours and the remainder over the next 24, with an added 10 mL/kg bolus. Its neurologic outcomes matched the slower arm.
Rapid rehydration was also associated with earlier normalization of the anion gap. Among children with severe acidosis, point estimates favored faster recovery of short-term memory scores rather than harm.
The trial defined its primary outcome as two consecutive Glasgow Coma Scale scores below 14 during treatment, a marker chosen because abnormal scores correlate with subtle cerebral edema on imaging. Results held when the analysis was restricted to children treated strictly per protocol. Memory and IQ measured two to six months after recovery were also similar across all four regimens.
Fluid tonicity: 0.9% versus 0.45% saline
PECARN FLUID also compared 0.9% and 0.45% saline directly. Neither concentration changed neurologic outcomes (RR 0.80, 95% CI 0.46–1.40 for GCS decline).
The concentrations differed only in non-neurologic effects. Isotonic 0.9% saline was associated with more hyperchloremic acidosis, hypocalcemia, and hypophosphatemia. Whether balanced crystalloids reduce these effects, or the risk of acute kidney injury, remains unsettled in children.
The balanced-crystalloid question is the live one for pediatric DKA management. A pediatric trial of roughly 77 children comparing a balanced solution with 0.9% saline found no difference in time to DKA resolution. Whether balanced fluids reduce hyperchloremia or kidney injury at scale awaits larger trials.
How large a deficit to assume
Assumed dehydration once ranged from 10% to 15% of body weight, but body-water studies later placed the median measured deficit nearer 6% to 9%. Bedside signs such as skin turgor and capillary refill proved unreliable in ketoacidosis, partly because hyperosmolarity preserves intravascular volume. Contemporary pediatric DKA management therefore favors a conservative assumed deficit and subtracts only boluses beyond 20 mL/kg from the calculated total.
Comparative Analysis
Weighing these studies means weighing their designs. The PECARN FLUID trial sits at the top: it randomized both rehydration rate and saline content, removing the confounding that clouds observational work, and found no effect of either on neurologic outcomes.
Two observational studies agree with it. Glaser and colleagues (2001) found no association between cerebral edema and fluid infusion rate (RR 1.1, 95% CI 0.4–3.0), sodium rate, or insulin rate. Lawrence and colleagues (2005) likewise found no association with any treatment factor.
One study points the other way. The UK case–control by Edge and colleagues (2006) linked cerebral edema to early insulin (OR 4.7, 95% CI 1.5–13.9) and to larger fluid volumes given in the first four hours.
What the biochemical risk factors share is severity, not treatment. Low PCO₂ (RR 3.4, 95% CI 1.9–6.3) and high blood urea predict injury across studies. Both mark hypoperfusion and dehydration, which fits the ischemia model rather than the osmotic-fluid model.
The subgroup findings reinforce this reading. In children presenting with the most severe acidosis or hypocapnia, faster rehydration was associated with quicker recovery of short-term memory, not with harm. Across per-protocol and as-treated analyses, no fluid regimen emerged as neurologically superior in pediatric DKA management.
Quality and Consistency of the Evidence
Confidence differs by question, and separating the well-answered from the open questions is central to sound pediatric DKA management.
- High — Rehydration rate does not meaningfully change neurologic outcomes. One large randomized trial with concordant observational data. Strong evidence supports this conclusion.
- High to moderate — Saline tonicity (0.9% vs 0.45%) does not change neurologic outcomes. Randomized evidence, internally consistent across arms.
- Moderate — Illness-severity markers predict cerebral injury. Low PCO₂, high blood urea, and bicarbonate treatment show consistent observational associations.
- Low — The single best absolute rate or deficit. Values outside the boundaries PECARN tested have not been randomized.
- Low / emerging — Balanced crystalloids versus 0.9% saline. Pediatric data on kidney injury and DKA resolution remain limited.
- Moderate — Ketoacidosis itself may affect later cognition. Cohorts link DKA, especially at diabetes onset, to subtle differences in memory and brain structure, independent of fluid strategy.
What the Evidence Does Not Show
The evidence is strong within its limits, and those limits matter. PECARN FLUID compared the upper and lower bounds of protocols already in use; rates far outside that range remain untested.
Because clinically apparent brain injury affects under 1% of episodes, no single trial can be powered to detect small differences in that outcome directly. Decline in mental status served as the practical surrogate.
The trial does not identify one optimal rate, tonicity, or deficit for every child. It argues against strict fluid restriction, not for a single universal regimen.
Two adjacent questions also stay open. Whether balanced crystalloids outperform 0.9% saline for kidney injury and acidosis in children is under-studied, and the optimal cadence of neurologic monitoring during the highest-risk early hours is not defined by trial data. Both are reasonable priorities for pediatric DKA management research.
Finally, the mechanism of injury remains incompletely mapped. A contribution from fluids in rare, extreme cases cannot be fully excluded — only shown to be minor across the range studied.
Practical Implications
The evidence supports a shift in pediatric DKA management from fluid restriction toward restoring perfusion. Five considerations follow, framed as what the data support rather than as directives.
1. Restore perfusion first
Adequate fluid to restore and maintain tissue perfusion is the priority, and current guidelines state this explicitly. An initial isotonic bolus — commonly 10–20 mL/kg over 20–30 minutes — is appropriate when perfusion is compromised, rather than withholding volume out of edema fear.
2. A rate within protocol bounds is safe
Both the faster and slower schedules studied produced equivalent neurologic outcomes, so the choice can follow severity and osmolality. Faster correction was associated with quicker metabolic recovery without added neurologic risk. Neither speed within the tested range should be avoided out of concern for cerebral injury.
3. Either tonicity is acceptable neurologically — watch electrolytes
0.9% and 0.45% saline gave the same neurologic results. Because 0.9% saline was associated with more hyperchloremic acidosis and electrolyte disturbance, monitor closely and let serum sodium and osmolality guide any transition.
4. Target the real risk factors with vigilance
Severe acidosis, marked hypocapnia, high blood urea, young age, and new-onset diabetes mark the highest-risk children. Monitor neurologic status closely, especially in the first hours, and recognize that injury can precede treatment. Deterioration clusters in the early treatment window, so that period warrants the closest attention.
5. Avoid routine bicarbonate; keep osmotic therapy ready
Bicarbonate treatment was associated with cerebral edema (RR 4.2, 95% CI 1.5–12.1) and is not recommended routinely. Keep mannitol or hypertonic saline immediately available, and treat suspected cerebral injury without delay once hypoglycemia is excluded.
Evidence Grade and Bottom Line
- Strong evidence supports that neither rehydration rate nor saline tonicity, within the ranges studied, meaningfully alters neurologic outcomes in children with DKA.
- Cerebral injury tracks illness severity — hypocapnia, high blood urea, and severe acidosis — and can begin before treatment, which implicates hypoperfusion and inflammation.
- Contemporary pediatric DKA management centers on restoring perfusion, individualizing correction, avoiding routine bicarbonate, monitoring closely, and keeping osmotic therapy at hand.
Article Information and References
- Kuppermann N, Ghetti S, Schunk JE, et al. Clinical trial of fluid infusion rates for pediatric diabetic ketoacidosis. N Engl J Med. 2018;378(24):2275–2287. DOI: https://doi.org/10.1056/NEJMoa1716816
- Glaser N, Fritsch M, Priyambada L, et al. ISPAD Clinical Practice Consensus Guidelines 2022: Diabetic ketoacidosis and hyperglycemic hyperosmolar state. Pediatr Diabetes. 2022;23(7):835–856. DOI: https://doi.org/10.1111/pedi.13406
- Glaser N, Barnett P, McCaslin I, et al. Risk factors for cerebral edema in children with diabetic ketoacidosis. N Engl J Med. 2001;344(4):264–269. DOI: https://doi.org/10.1056/NEJM200101253440404
- Lawrence SE, Cummings EA, Gaboury I, Daneman D. Population-based study of incidence and risk factors for cerebral edema in pediatric diabetic ketoacidosis. J Pediatr. 2005;146(5):688–692. DOI: https://doi.org/10.1016/j.jpeds.2004.12.041
- Edge JA, Jakes RW, Roy Y, et al. The UK case–control study of cerebral oedema complicating diabetic ketoacidosis in children. Diabetologia. 2006;49(9):2002–2009. DOI: https://doi.org/10.1007/s00125-006-0363-8
- Glaser NS, Ghetti S, Casper TC, Dean JM, Kuppermann N. Pediatric diabetic ketoacidosis, fluid therapy, and cerebral injury: the design of a factorial randomized controlled trial. Pediatr Diabetes. 2013;14(6):435–446. DOI: https://doi.org/10.1111/pedi.12027
- Glaser NS, Wootton-Gorges SL, Marcin JP, et al. Mechanism of cerebral edema in children with diabetic ketoacidosis. J Pediatr. 2004;145(2):164–171. DOI: https://doi.org/10.1016/j.jpeds.2004.03.045
- Tasker RC, Acerini CL. Cerebral edema in children with diabetic ketoacidosis: vasogenic rather than cellular? Pediatr Diabetes. 2014;15(4):261–270. DOI: https://doi.org/10.1111/pedi.12153
- Hsia DS, Tarai SG, Alimi A, Coss-Bu JA, Haymond MW. Fluid management in pediatric patients with diabetic ketoacidosis and rates of suspected clinical cerebral edema. Pediatr Diabetes. 2015;16(5):338–344. DOI: https://doi.org/10.1111/pedi.12268
- Long B, Koyfman A. Emergency medicine myths: cerebral edema in pediatric diabetic ketoacidosis and intravenous fluids. J Emerg Med. 2017;53(2):212–221. DOI: https://doi.org/10.1016/j.jemermed.2017.03.014
- Jayashree M, Williams V, Iyer R. Fluid therapy for pediatric patients with diabetic ketoacidosis: current perspectives. Diabetes Metab Syndr Obes. 2019;12:2355–2361. DOI: https://doi.org/10.2147/DMSO.S194944
- Yung M, Letton G, Keeley S. Controlled trial of Hartmann’s solution versus 0.9% saline for diabetic ketoacidosis. J Paediatr Child Health. 2017;53(1):12–17. DOI: https://doi.org/10.1111/jpc.13436