Clinical question
In children presenting to an emergency department with suspected septic shock and abnormal perfusion, does using a balanced crystalloid rather than 0.9% saline reduce death or major adverse kidney events?
Paediatric sepsis guidelines have generally favoured balanced crystalloids because saline contains a supraphysiological chloride concentration and may cause:
The perceived clinical benefit comes from physiological reasoning, observational studies and adult trials. Does the use of balanced fluids improve meaningful outcomes in children?
PRoMPT BOLUS is the largest randomised trial addressing this question in paediatric septic shock.
"We hypothesized that resuscitation and maintenance hydration with balanced fluid would lead to a lower incidence of major adverse kidney events within 30 days than that with 0.9% saline."
THE STUDY
Balamuth F, Weiss SL, Long E, et al.; PRoMPT BOLUS Investigators.
New England Journal of Medicine. 2026;395:870–881.
What They Did
This was a large, pragmatic, multicentre, open-label, randomised controlled trial in 47 emergency departments across United States, Canada, Australia, New Zealand and Costa Rica
Population
N = 8482
Children aged 2 months to under 18 years with:
-
Suspected septic shock
-
Clinical evidence of abnormal perfusion
-
A requirement for crystalloid fluid resuscitation
-
No more than 40 mL/kg of crystalloid administered before enrolment
The pragmatic eligibility criteria were intended to identify children early, without waiting for laboratory confirmation of shock or organ dysfunction.
Intervention
Balanced crystalloid, including:
The allocated fluid was preferentially used for:
Treatment continued until 11:59 pm on the day following enrolment, producing an intervention period of approximately 24–48 hours.
Comparison Fluid
0.9% sodium chloride used for the same purposes and duration.
The timing, rate and total volume of fluid remained at the treating clinician’s discretion.
Primary outcome
A major adverse kidney event within 30 days, defined as a composite of:
Persistent kidney dysfunction required a final serum creatinine at least 200% of baseline, with a minimum increase of 0.3 mg/dL.
The endpoint was assessed at 30 days or hospital discharge, whichever occurred first.
Study population
A total of 9,041 children were randomised. Baseline characteristics were well balanced. After withdrawal or refusal of permission to use data:
-
Balanced fluid: 4,235 patients
-
0.9% saline: 4,247 patients
Important features of the cohort included:
| Characteristic |
Finding |
| Median age |
6.8 years |
| Median initial lactate |
1.9 mmol/L |
| Respiratory source of infection |
46.6% |
| Any acute kidney injury at enrolment |
23.8% |
| Vasoactive medication |
14.2% |
| Invasive ventilation |
9.7% |
| Hospital mortality |
1.1% |
Low lactate, low use of vasoactive support and low mortality, suggest that many participants had relatively mild or early septic shock rather than severe refractory shock.
Fluid exposure
The groups received similar total fluid volumes.
| Fluid exposure |
Balanced group |
Saline group |
| Median total crystalloid volume |
85 mL/kg |
88 mL/kg |
| Received ≥75% allocated fluid |
80% |
88% |
| Median saline exposure |
20 mL/kg |
79 mL/kg |
| Median balanced-fluid exposure |
58 mL/kg |
0 mL/kg |
There was meaningful separation between groups, although the balanced-fluid group still received a median of 20 mL/kg of saline, much of it before randomisation.
Main results
Primary outcome
| Major adverse kidney event |
Balanced fluid |
0.9% saline |
| Event rate |
137/4,073 (3.4%) |
124/4,068 (3.0%) |
| Absolute difference |
+0.4 percentage points |
|
| Risk ratio |
1.10 |
|
| 95% CI |
0.88–1.40 |
|
| P value |
0.85 |
|
Balanced fluid did not reduce the composite of death, renal-replacement therapy or persistent kidney dysfunction.
The point estimate slightly favoured saline, but the confidence interval crossed 1 and included both modest benefit and potential harm from balanced fluid.
Components of the primary outcome
| Outcome |
Balanced fluid |
0.9% saline |
Risk ratio |
| Death within 30 days |
1.0% |
0.9% |
1.07 |
| New renal-replacement therapy |
0.6% |
0.7% |
0.84 |
| Persistent kidney dysfunction |
2.3% |
1.9% |
1.15 |
| Death before discharge |
1.1% |
1.1% |
1.02 |
| Death within 90 days |
2.3% |
2.1% |
1.07 |
None of these outcomes demonstrated a clinical difference.
Hospital outcomes
Electrolyte and biochemical effects
| Safety outcome |
Balanced fluid |
0.9% saline |
| Hyperchloraemia, chloride >110 mmol/L |
31.4% |
49.0% |
| Hypernatraemia, sodium >155 mmol/L |
1.8% |
3.1% |
| Hyperlactataemia, lactate >4 mmol/L |
19.8% |
16.7% |
| Thrombosis |
1.3% |
1.3% |
| Cerebral oedema |
0.4% |
0.4% |
Balanced fluid produced a large absolute reduction in hyperchloraemia of approximately 17.6 percentage points, but this biochemical advantage did not translate into fewer kidney events, shorter hospitalisation or improved survival.
Subgroup findings
No convincing treatment effect was identified according to:
Some estimates numerically favoured balanced fluid among children receiving the largest fluid volumes or presenting with more marked acidosis or hyperlactataemia, but these analyses were not definitive.
The subgroup aged 12–17 years numerically favoured saline, but isolated subgroup results should not override the neutral overall trial, particularly without a consistent biological pattern or correction for multiple comparisons.
Strengths
1. Large sample size
With almost 9,000 participants, this is by far the most substantial paediatric randomised trial of balanced crystalloid versus saline in suspected septic shock.
2. Clinically relevant setting
Children were enrolled early in the emergency department, when fluid selection decisions are actually being made.
3. International generalisability
The trial included hospitals across five countries and several major paediatric emergency research networks.
4. Pragmatic intervention
Clinicians retained control over the timing and volume of fluids, making the findings applicable to routine emergency care.
5. Meaningful treatment separation
Although there was some crossover and pre-randomisation saline use, the two groups had clearly different crystalloid exposure and different rates of hyperchloraemia.
6. Patient-centred primary outcome
The primary endpoint focused on death, renal-replacement therapy and persistent renal dysfunction, rather than a transient laboratory change.
Limitations
1. The population had a low event rate
The trial assumed a primary-event rate of approximately 6% in the saline group, but the observed rate was only 3%. This reduced the effective statistical power. The study confidently excludes the large benefit originally anticipated, but it cannot exclude a smaller clinically important advantage—or disadvantage—of either fluid.
A neutral superiority trial is not proof that the fluids are completely equivalent.
2. Many patients were not severely shocked
Only 14.2% received vasoactive therapy, fewer than 10% were invasively ventilated, median lactate was 1.9 mmol/L and mortality was 1.1%.
Therefore, the findings primarily apply to children treated early for suspected or relatively moderate septic shock, rather than those with profound shock, severe acidosis, multiorgan failure or massive fluid requirements.
3. Open-label design
Clinicians knew the allocated fluid. This was pragmatically necessary but could have influenced subsequent fluid choice, laboratory testing or other treatments.
4. Incomplete laboratory testing
A measured baseline creatinine was unavailable in approximately 40% of patients. Age- and sex-based estimates were used, potentially introducing some misclassification of kidney dysfunction.
Electrolyte and lactate outcomes were based only on children who underwent follow-up testing. Testing was not protocolised for all participants and may have been influenced by illness severity.
5. Intervention limited to 24–48 hours
Fluid administered after the intervention period was not controlled and may have affected later outcomes.
6. Limited applicability to other settings
The results may not extend to:
-
Low-resource settings
-
Hospital-acquired septic shock
-
Neonates younger than two months
-
Children with major burns or traumatic shock
-
Profound refractory septic shock
-
Patients requiring fluid compositions tailored to specific electrolyte disorders
Interpretation
PRoMPT BOLUS challenges the assumption that the physiological advantages of balanced crystalloids necessarily produce better clinical outcomes in children.
Balanced fluids clearly caused less hyperchloraemia and hypernatraemia. However, within this relatively low-mortality paediatric population, those biochemical improvements did not reduce:
The study does not demonstrate that saline is superior. Nor does it prove absolute equivalence. It demonstrates that routine preferential use of balanced fluid did not produce the hypothesised large improvement in major kidney outcomes.
For most children receiving fluid resuscitation for suspected septic shock:
-
Either balanced crystalloid or 0.9% saline is a reasonable initial choice. There is no evidence from this trial that changing every paediatric septic-shock bolus from saline to balanced fluid improves survival or prevents major kidney injury.
-
Balanced crystalloid remains attractive when hyperchloraemia, hypernatraemia or metabolic acidosis is a concern.
-
Saline should not be considered harmful solely because it produces more hyperchloraemia.
-
Fluid volume, repeated assessment and avoidance of fluid overload are probably more important than choosing between these two crystalloid categories.
-
The findings should not be extrapolated too confidently to profoundly shocked children receiving very large-volume resuscitation.
Literature Rounds verdict
Practice informing, but not necessarily practice changing.
Balanced crystalloids reduce hyperchloraemia, but PRoMPT BOLUS found no corresponding improvement in major kidney events, survival or hospital-free days. In routine paediatric emergency sepsis resuscitation, clinicians can reasonably use either balanced crystalloid or 0.9% saline, guided by the child’s electrolyte profile, acid–base state, comorbidities and local availability. Fluid selection appears less important than careful dosing and repeated reassessment.
