Publication

• Title: Balanced Fluid or 0.9% Saline in Children Treated for Septic Shock
• Acronym: PRoMPT BOLUS
• Year: 2026
• Journal published in: New England Journal of Medicine
• Citation: Balamuth F, Weiss SL, Long E, et al. Balanced fluid or 0.9% saline in children treated for septic shock. N Engl J Med. Published online April 24, 2026.

Context & Rationale

• Background

  • Crystalloid fluid is the default initial resuscitation fluid for children with suspected septic shock.
  • 0.9% saline is isotonic but chloride-rich, with sodium and chloride concentrations of 154 mEq/L.
  • Balanced crystalloids such as lactated Ringer’s, Plasma-Lyte, and Hartmann’s solution contain lower chloride and buffer anions, more closely approximating plasma composition.
  • The biological argument for balanced fluid is plausible: less hyperchloraemia, less metabolic acidosis, less renal vasoconstriction, and less kidney injury.
  • The clinical question remained unsettled because biochemical effects had not reliably translated into patient-centred benefit in children.
  • The PRoMPT BOLUS protocol framed this as a pragmatic kidney-centred comparative-effectiveness question using major adverse kidney events within 30 days as the primary outcome.¹
  • A preceding pilot feasibility study showed that emergency-department enrolment, treatment separation, and enrolment under emergency consent processes were operationally feasible.²
• Research Question/Hypothesis

  • The trial tested whether balanced crystalloids, used for bolus resuscitation and maintenance hydration for up to 48 hours, would reduce death, new renal-replacement therapy, or persistent kidney dysfunction compared with 0.9% saline in children treated for suspected septic shock.
  • The trialists hypothesised that balanced fluid would reduce major adverse kidney events within 30 days.
• Why This Matters

  • The intervention is cheap, immediately available, and globally familiar.
  • A small absolute treatment effect would matter because crystalloids are administered to large numbers of children with suspected sepsis.
  • The 2026 Surviving Sepsis Campaign guideline suggested balanced or buffered crystalloids over 0.9% saline for children with septic shock requiring fluid boluses, but rated the certainty of evidence as very low and stated that 0.9% saline remained a suitable alternative.³
  • PRoMPT BOLUS was therefore positioned to test whether a guideline-preferred but weakly supported fluid strategy improved outcomes in real-world paediatric emergency care.

Design & Methods

• Research Question: In children treated for suspected septic shock in the emergency department, does resuscitation and maintenance hydration with balanced crystalloid fluid reduce major adverse kidney events within 30 days compared with 0.9% saline?
• Study Type: Randomised, multicentre, pragmatic, open-label, interventional, patient-level, site-stratified, international trial conducted in 47 emergency departments across the United States, Canada, Australia, New Zealand, and Costa Rica.
• Population:
  • Final eligible age range was 2 months to <18 years.
  • Patients had suspected septic shock and were being treated with parenteral antibiotics and crystalloid fluid resuscitation for abnormal perfusion.
  • Abnormal perfusion was determined clinically, including hypotension or abnormal capillary refill.
  • Total crystalloid volume before enrolment had to be confirmed as ≤40 mL/kg.
  • Patients could be enrolled after an emergency-department sepsis alert with clinician confirmation or clinician diagnosis/treatment for septic shock.
  • Exclusions included clinician judgement that either fluid would be unsafe, suspicion for impending brain herniation, known severe hyperkalaemia, known severe hypercalcaemia, known fulminant hepatic failure, dialysis-dependent kidney disease, severe hepatic impairment, metabolic or mitochondrial disorders, primary mineralocorticoid deficiency, known pregnancy, known prisoner status, known allergy to either fluid, prior opt-out, or inability to confirm pre-randomisation crystalloid volume ≤40 mL/kg.
  • Enrolment used exception from informed consent, deferred consent, delayed consent, independent medical practitioner consent, or prospective consent according to local regulations.
• Intervention:
  • Balanced crystalloid fluid was preferentially used for all subsequent crystalloid boluses and as the base maintenance fluid until 11:59 p.m. on the calendar day after randomisation.
  • Balanced crystalloids included lactated Ringer’s solution, Plasma-Lyte, or Hartmann’s solution according to local availability and clinician preference.
  • The intervention window ensured approximately 24 to 48 hours of assigned study-fluid exposure.
  • Clinicians determined the timing, rate, dose, total volume, and need for further fluid, vasoactive medication, mechanical ventilation, antibiotics, corticosteroids, buffers, and all other care.
  • Alternative crystalloid fluids could be used for clinical indications such as electrolyte abnormalities or safety concerns.
• Comparison:
  • 0.9% saline was preferentially used for all subsequent crystalloid boluses and as the base maintenance fluid through the same intervention window.
  • The same clinician-directed approach to timing, volume, rate, and co-interventions applied in the saline group.
  • Hospital fluid supplies were used without study-specific labelling.
• Blinding: Patients and clinicians were not blinded because the fluids were embedded into emergency care and electrolyte changes could reveal treatment allocation; the senior biostatistician and investigators remained unaware of aggregate outcomes until enrolment was complete.
• Statistics: A total of 8800 patients was required to detect a 1.7 percentage-point absolute reduction in major adverse kidney events within 30 days, from 6.0% with 0.9% saline to 4.3% with balanced fluid, with 95% power and a two-sided type I error of 0.05; the final target was increased to 9178 to account for anticipated withdrawal. The primary analysis was an analysed-as-randomised, consent-limited modified intention-to-treat analysis excluding patients whose legally authorised representative withdrew or declined data use; multiple imputation was used for missing outcome components, ineligible patients identified after randomisation were retained, and complete-case and per-protocol analyses were prespecified sensitivity analyses.
• Follow-Up Period: The primary outcome was assessed at 30 days after enrolment or hospital discharge, whichever occurred first; hospital-free days were assessed through 28 days, laboratory safety outcomes through 4 days, adverse events through 7 days, and mortality through 90 days.

Key Results

This trial was not stopped early for efficacy, futility, or safety. Three interim analyses were reviewed by the DSMB at 15%, 40%, and 70% enrolment, and the DSMB recommended continuation without modification; final enrolment was 9041 patients, slightly below the 9178 target.

• The primary result was decisively neutral for the overall population: MAKE30 was 3.4% with balanced fluid versus 3.0% with saline, RR 1.10; 95% CI 0.88 to 1.40; P=0.85.
• The trial demonstrated strong biochemical separation but not clinical separation: hyperchloraemia fell from 49.0% to 31.4%, yet death, renal-replacement therapy, persistent kidney dysfunction, hospital stay, and hospital-free days were unchanged.
• Subgroup analyses did not identify a reliable treatment effect; examples include total crystalloid >100 mL/kg, 48/1588 (3.0%) versus 62/1674 (3.7%), RR 0.84; 95% CI 0.59 to 1.20, and any AKI at enrolment, 95/962 (9.9%) versus 83/1052 (7.9%), RR 1.29; 95% CI 0.98 to 1.71.

Internal Validity

• Randomisation and Allocation: Randomisation used 1:1 permuted blocks stratified by site, with allocation concealed in sequentially numbered opaque envelopes opened only after eligibility confirmation.
• Dropout or Exclusions: Of 9041 randomised patients, 559 were excluded from analysis because data use was not permitted: 277/4512 (6.1%) in the balanced-fluid group and 282/4529 (6.2%) in the saline group.
• Missing Outcomes: In the analysed population, some MAKE30 components were missing for 162/4235 balanced-fluid patients and 179/4247 saline patients, triggering multiple imputation.
• Robustness to Missingness: Complete-case analysis was similar, with MAKE30 RR 1.1; 95% CI 0.87 to 1.4, and per-protocol analysis was also similar, with MAKE30 RR 1.1; 95% CI 0.80 to 1.5.
• Performance/Detection Bias: The trial was open-label, so treating clinicians knew the assigned fluid; however, the main outcomes were largely objective: death, renal-replacement therapy, and creatinine-defined persistent kidney dysfunction.
• Laboratory Measurement Bias: Safety outcomes such as lactate, chloride, sodium, potassium, calcium, and creatinine depended on clinician-ordered laboratory testing rather than mandatory scheduled testing, leaving potential differential ascertainment for laboratory-defined harms.
• Protocol Adherence: Among patients receiving any post-randomisation crystalloid, ≥75% assigned-fluid adherence occurred in 3300/4128 (80%) in the balanced-fluid group and 3661/4147 (88%) in the saline group.
• Baseline Characteristics: Groups were well balanced: median age was 6.8 years in both groups, male sex was 51.2% versus 50.4%, cancer was 14.3% versus 14.5%, and baseline creatinine was 0.30 mg/dL in both groups.
• Illness Severity: The cohort included substantial comorbidity and early organ risk, but overall severity was modest by critical-care standards: vasoactive medications were used in 14.2%, invasive mechanical ventilation in 9.7%, hospital mortality was 1.1%, and MAKE30 in the saline arm was 3.0% rather than the assumed 6.0%.
• Heterogeneity: Clinical heterogeneity was intentional and appropriate for a pragmatic trial; the trade-off is that community-acquired suspected septic shock, heterogeneous balanced fluids, and clinician-directed resuscitation may dilute a treatment effect in a narrower biological phenotype.
• Timing: Enrolment was early relative to hospital care but not before the first fluid exposure; median time from emergency department arrival to enrolment was 103 minutes with balanced fluid and 101 minutes with saline, and median pre-enrolment 0.9% saline volume was 18 versus 17 mL/kg.
• Dose: The trial tested fluid composition, not a fixed dose or resuscitation-volume strategy; total crystalloid exposure was similar at 85 mL/kg (IQR 55 to 119) versus 88 mL/kg (IQR 57 to 123), median difference −2.3 mL/kg; 95% CI −4.3 to −0.30.
• Separation of the Variable of Interest: Median 0.9% saline exposure was 20 mL/kg (IQR 5.5 to 32) in the balanced-fluid group versus 79 mL/kg (IQR 49 to 113) in the saline group, median difference −57 mL/kg; median balanced-fluid exposure was 58 mL/kg (IQR 31 to 92) versus 0 mL/kg (IQR 0 to 0), median difference +54 mL/kg.
• Key Delivery Aspects: Maintenance fluid contributed materially to total fluid delivery: median maintenance crystalloid volume was 45 mL/kg versus 47 mL/kg, and balanced maintenance fluid was 35 mL/kg versus 0 mL/kg.
• Crossover: Cross-fluid exposure was present but did not eliminate separation; the lower adherence in the balanced-fluid arm could dilute a true benefit, but the per-protocol analysis did not show a different clinical effect.
• Adjunctive Therapy Use: Co-interventions were similar: vasoactive medications 14% versus 14%, invasive mechanical ventilation 9.7% versus 9.7%, corticosteroids 23% versus 22%; bicarbonate or other buffer use was lower with balanced fluid, 4.6% versus 5.9%.
• Outcome Assessment: The primary composite was clinically meaningful but internally heterogeneous: death, dialysis, and creatinine-defined persistent kidney dysfunction are not equivalent outcomes, and persistent kidney dysfunction was the numerically dominant component.
• Baseline Creatinine: Baseline creatinine was measured in 2515 balanced-fluid patients and 2529 saline patients, but imputed in 1717 and 1714 respectively, which is unavoidable in paediatric emergency research but relevant for creatinine-based outcomes.
• Statistical Rigor: The analysis matched the design, adjusted for site, used a prespecified final alpha threshold of P<0.044 after interim looks, retained post-randomisation ineligible patients in the primary analysis, and used multiple imputation plus complete-case, per-protocol, and tipping-point sensitivity analyses.
• Power Actually Achieved: The lower-than-expected primary event rate reduced practical power to detect the planned absolute benefit, especially for mortality, dialysis, and severe-illness subgroups.

Conclusion on Internal Validity: Internal validity is strong for the main overall null conclusion because randomisation was concealed, the trial was large, outcomes were objective, treatment separation was real, and sensitivity analyses were consistent. It is more limited for rare harms, laboratory-defined safety outcomes, and severe biological subgroups because of open-label care, clinician-directed testing, missing laboratory data, lower event rates, and limited subgroup power.

External Validity

• Population Representativeness: Participants were typical of children treated for suspected community-acquired septic shock in modern paediatric emergency departments with intensive-care access.
• Geography: Enrolment was broad but predominantly high-resource: 77% of analysed patients were enrolled in the United States, 8.5–8.7% in Canada, 14% in Australia or New Zealand, and approximately 0.65% in Costa Rica.
• Clinical Spectrum: The trial captured real emergency-department uncertainty: 46–47% had pneumonia or other respiratory infection, 8–9% bloodstream infection, 8–10% abdominal infection, 8–9% genitourinary infection, 12% versus 11% had any positive blood culture, and 5.8% versus 5.7% were later judged to have an alternative diagnosis rather than sepsis.
• Applicability to Current Practice: The pragmatic design, clinician-directed fluid volume, and use of hospital stock fluids make the findings highly applicable to emergency departments where both balanced fluids and saline are available.
• Limited Applicability: The findings do not directly generalise to neonates younger than 2 months, children with hospital-acquired sepsis, children with severe pre-existing renal or hepatic disease, metabolic disorders, impending brain herniation, resource-limited settings without intensive-care availability, or settings where sepsis epidemiology is dominated by malaria, dengue shock, severe anaemia, or malnutrition.
• Fluid Availability: The results are especially useful in systems where balanced fluids and saline are both acceptable but local logistics, cost, pharmacy stock, or compatibility issues influence fluid choice.
• Patient-Specific Exceptions: The trial supports broad equivalence, not indiscriminate interchangeability; saline may still be preferred for particular electrolyte or neurological concerns, while balanced fluid may remain attractive when avoiding hyperchloraemia is a priority.

Conclusion on External Validity: External validity is strong for high-resource paediatric emergency departments treating suspected community-acquired septic shock. Generalisability is limited for neonates, low-resource settings, hospital-acquired sepsis, selected metabolic/renal/hepatic disorders, and children with intracranial pathology or marked electrolyte constraints.

Strengths & Limitations

• Strengths:
  • Very large paediatric emergency-department trial in a field where most prior evidence was adult, observational, biochemical, or underpowered.
  • International multicentre design across 47 sites and five countries.
  • Early enrolment before large-volume resuscitation, with median pre-enrolment crystalloid approximately 17–18 mL/kg.
  • Clear treatment separation in delivered fluid type and biochemical chloride/sodium effects.
  • Objective primary-outcome components and robust sensitivity analyses.
  • Pragmatic conduct, making the result directly relevant to routine care.
  • Ethically sophisticated emergency-research consent framework.
• Limitations:
  • Open-label treatment allocation.
  • Withdrawal or declined data use in approximately 6% of randomised patients.
  • Primary outcome event rate substantially lower than expected, reducing power for the planned absolute effect and severe subgroups.
  • Creatinine-defined outcome partly dependent on imputed baseline creatinine and clinician-ordered follow-up creatinine.
  • Composite primary outcome combines death, renal-replacement therapy, and persistent kidney dysfunction, which differ in patient importance.
  • Laboratory safety outcomes had missingness because tests were ordered as part of clinical care rather than mandated by trial schedule.
  • Only 80% of balanced-fluid patients receiving post-randomisation crystalloid met the ≥75% assigned-fluid adherence threshold.
  • Balanced-fluid arm combined lactated Ringer’s, Hartmann’s solution, and Plasma-Lyte rather than testing one specific crystalloid.
  • The trial did not test restrictive versus liberal fluid volume, early vasoactive timing, or haemodynamic resuscitation targets.
  • Biomarker substudy aims described in the protocol were not reported in the index manuscript.

Interpretation & Why It Matters

• Clinical Message

  For most children treated for suspected septic shock in high-resource emergency departments, balanced crystalloid fluid did not improve kidney-centred outcomes, mortality, renal-replacement therapy, hospital length of stay, or hospital-free days compared with 0.9% saline.
• Biological Message

  Balanced fluids clearly reduced hyperchloraemia and hypernatraemia, but those biochemical benefits did not translate into the patient-centred outcomes chosen for this trial.
• Practice Implication

  Fluid choice can reasonably be guided by local availability, compatibility, clinician workflow, and patient-specific electrolyte or neurological considerations rather than a presumed overall survival or kidney advantage for balanced crystalloids.
• Guideline Implication

  The trial materially weakens the case for a universal preference for balanced fluids in paediatric septic shock, although it does not prove equivalence in every severe subgroup or every clinical context.
• Research Implication

  The next important questions are likely phenotype-specific: children with very high fluid exposure, severe acidosis, severe AKI at presentation, fluid-refractory shock, or settings where chloride load is extreme.

Controversies & Other Evidence

• Adult extrapolation was always uncertain: The enthusiasm for balanced crystalloids was heavily influenced by SALT-ED and SMART, which showed small kidney-composite advantages in adults, while an accompanying editorial cautioned that widespread practice change should be anchored in patient-centred validation across settings and populations.⁴⁵⁶
• The adult critical-care evidence became more mixed: BaSICS, PLUS, and FLUID did not demonstrate decisive patient-centred superiority for balanced fluids across broad critically ill or hospital-wide populations, while adult meta-analysis has continued to suggest possible but not definitive average benefit.⁷⁸⁹¹⁰
• Paediatric biochemical benefit preceded clinical uncertainty: SPLYT-P showed that balanced solutions reduced chloride rise in paediatric intensive-care patients, reinforcing the biological plausibility of balanced fluids without establishing a major clinical-outcome advantage.¹¹
• The smaller paediatric septic-shock signal was not reproduced: Sankar and colleagues reported less new or progressive AKI with a multiple-electrolyte solution in 708 children, and contemporaneous commentary interpreted that trial as important support for balanced bolus fluids; PRoMPT BOLUS was much larger, used a stricter 30-day kidney composite, and did not reproduce a patient-centred kidney benefit.¹²¹³
• Pre-PRoMPT paediatric meta-analyses now require reinterpretation: Recent paediatric systematic reviews before the index result suggested no mortality advantage but possible reductions in AKI, renal-replacement therapy, and hyperchloraemia with balanced fluids; PRoMPT BOLUS contributes many more patients than the prior RCT evidence base and shifts the centre of gravity away from routine patient-centred superiority.¹⁴¹⁵
• The 2026 guideline preference should be reappraised: The Surviving Sepsis Campaign recommendation favouring balanced or buffered crystalloids in children was conditional and based on very low certainty; PRoMPT BOLUS supports retaining 0.9% saline as an acceptable standard option and reduces the strength of any blanket preference for balanced fluids.³
• Event-rate dilution is central to interpretation: MAKE30 occurred in only 3.0% of saline patients, approximately half the planned control rate, and discharge mortality was only 1.1%; this supports broad safety but limits inference for the sickest children.
• Outcome choice remains debatable: MAKE30 is objective and pragmatic, but it is a composite of clinically unequal components, and the persistent kidney dysfunction component depends heavily on creatinine measurement and baseline creatinine estimation.
• The trial answers composition, not resuscitation strategy: PRoMPT BOLUS does not adjudicate optimal bolus volume, restrictive fluid strategies, dynamic fluid responsiveness, or earlier vasoactive initiation.
• The negative result is not explained by lack of biochemical effect: Hyperchloraemia and hypernatraemia were materially lower with balanced fluid, so the central finding is that preventing these biochemical abnormalities did not improve the selected clinical outcomes in the overall trial population.

Summary

• PRoMPT BOLUS randomised 9041 children with suspected septic shock in 47 emergency departments to balanced fluid or 0.9% saline for bolus and maintenance crystalloid therapy for up to 48 hours.
• The primary outcome, major adverse kidney event within 30 days, occurred in 3.4% versus 3.0%; RR 1.10; 95% CI 0.88 to 1.40; P=0.85.
• Mortality, new renal-replacement therapy, persistent kidney dysfunction, hospital length of stay, and hospital-free days were not improved by balanced fluid.
• Balanced fluid reduced hyperchloraemia and hypernatraemia but increased hyperlactataemia slightly; thrombosis, cerebral oedema, and other safety outcomes were not meaningfully different.
• The trial is highly practice-informing for high-resource paediatric emergency departments, but it remains less definitive for neonates, low-resource settings, hospital-acquired sepsis, and the most severely ill shock phenotypes.

Overall Takeaway

PRoMPT BOLUS is a practice-shaping neutral trial: it shows that a balanced-fluid strategy reduces chloride and sodium abnormalities but does not improve kidney-centred or mortality outcomes in the overall population of children treated for suspected septic shock in high-resource emergency departments. It supports pragmatic equipoise between balanced crystalloid fluid and 0.9% saline for most such children, with fluid choice individualised to patient physiology, compatibility, and local availability.

Bibliography

• 1.Weiss SL, Balamuth F, Long E, Thompson G, Hayes KL, Katcoff H, et al. PRagMatic Pediatric Trial of Balanced vs nOrmaL Saline FlUid in Sepsis: study protocol for the PRoMPT BOLUS randomized interventional trial. Trials. 2021;22(1):776.
• 2.Balamuth F, Kittick M, McBride P, Woodford AL, Vestal N, Casper TC, et al. Pragmatic Pediatric Trial of Balanced Versus Normal Saline Fluid in Sepsis: the PRoMPT BOLUS randomized controlled trial pilot feasibility study. Acad Emerg Med. 2019;26(12):1346-1356.
• 3.Weiss SL, Peters MJ, Oczkowski SJW, et al. Surviving Sepsis Campaign international guidelines for the management of sepsis and septic shock in children 2026. Pediatr Crit Care Med. 2026;27(4):379-434.
• 4.Self WH, Semler MW, Wanderer JP, Wang L, Byrne DW, Collins SP, et al. Balanced crystalloids versus saline in noncritically ill adults. N Engl J Med. 2018;378(9):819-828.
• 5.Semler MW, Self WH, Wanderer JP, Ehrenfeld JM, Wang L, Byrne DW, et al. Balanced crystalloids versus saline in critically ill adults. N Engl J Med. 2018;378(9):829-839.
• 6.Myburgh J. Patient-centered outcomes and resuscitation fluids. N Engl J Med. 2018;378(9):862-863.
• 7.Zampieri FG, Machado FR, Biondi RS, Freitas FGR, Veiga VC, Figueiredo RC, et al. Effect of intravenous fluid treatment with a balanced solution vs 0.9% saline solution on mortality in critically ill patients: the BaSICS randomized clinical trial. JAMA. 2021;326(9):818-829.
• 8.Finfer S, Micallef S, Hammond N, Navarra L, Bellomo R, Billot L, et al. Balanced multielectrolyte solution versus saline in critically ill adults. N Engl J Med. 2022;386(9):815-826.
• 9.McIntyre L, Fergusson D, McArdle T, English S, Cook DJ, Fox-Robichaud AE, et al. A crossover trial of hospital-wide lactated Ringer’s solution versus normal saline. N Engl J Med. 2025;393(7):660-670.
• 10.Hammond NE, Zampieri FG, Di Tanna GL, Garside T, Adigbli D, Cavalcanti AB, et al. Balanced crystalloids versus saline in critically ill adults: a systematic review with meta-analysis. NEJM Evid. 2022;1(2):EVIDoa2100010.
• 11.Raman S, Gibbons KS, Mattke A, Newth CJL, Rickard CM, Festa MS, et al. Effect of saline vs gluconate/acetate-buffered solution vs lactate-buffered solution on serum chloride among children in the pediatric intensive care unit: the SPLYT-P randomized clinical trial. JAMA Pediatr. 2023;177(2):122-131.
• 12.Sankar J, Muralidharan J, Lalitha AV, Rameshkumar R, Pathak M, Das RR, et al. Multiple electrolytes solution versus saline as bolus fluid for resuscitation in pediatric septic shock: a multicenter randomized clinical trial. Crit Care Med. 2023;51(11):1449-1460.
• 13.Palmer ES, Agus MSD. Running wide open: strong evidence for bolus fluids selection in pediatric sepsis. Crit Care Med. 2023;51(11):1587-1589.
• 14.Chandelia S, Angurana SK, Prasad S, Tiwari MK, Jayashree M, Nallasamy K, et al. Balanced salt solution versus normal saline as resuscitation fluid in pediatric septic shock: a systematic review and meta-analysis of randomized controlled trials. J Intensive Care Med. 2026;41(4):274-283.
• 15.Vijendra B, Bertol AB, de Almeida MMG, de Freitas PHA, Simão AMS, de Faria BL. Balanced crystalloid versus saline for resuscitation in pediatric septic shock: a systematic review and meta-analysis. BMC Pediatr. 2025;25(1):81.