Publication

• Title: A crossover trial of hospital-wide lactated Ringer’s solution versus normal saline
• Acronym: FLUID
• Year: 2025
• Journal published in: New England Journal of Medicine
• Citation: McIntyre L, Fergusson D, McArdle T, et al. A crossover trial of hospital-wide lactated Ringer’s solution versus normal saline. N Engl J Med. 2025;393:660-670.

Context & Rationale

• Background

  • 0.9% saline remains a common default crystalloid across emergency, perioperative and critical care pathways.
  • Saline’s supraphysiological chloride load is biologically plausible as a driver of hyperchloraemic metabolic acidosis and renal vasoconstriction, and has been associated with adverse renal outcomes in non-randomised work.
  • Prior pragmatic and individually randomised trials comparing balanced crystalloids with saline suggested (at most) small absolute effects, leaving uncertainty about whether a whole-hospital policy change would translate into meaningful patient-centred benefits at scale.
  • Most prior definitive trials were ICU- or ED-centred; the “denominator” of general ward admissions and surgical/maternity/paediatric pathways remained comparatively under-tested in a policy-relevant way.
• Research Question/Hypothesis

  • Does implementing lactated Ringer’s as the default, hospital-wide intravenous crystalloid (versus normal saline) reduce the 90-day composite of death or hospital readmission among all hospital admissions?
• Why This Matters

  • Crystalloid choice is typically decided at formulary and supply-chain level; a pragmatic “default fluid” strategy could affect very large numbers of patients.
  • Even modest absolute benefits or harms could be important in population terms when multiplied across routine care volumes.
  • A hospital-wide trial tests implementation realism (stocking, order sets, clinical overrides) and external validity beyond highly selected ICU populations.

Design & Methods

• Research Question: Among patients admitted to hospital, does a hospital-wide policy of lactated Ringer’s as the default crystalloid (vs normal saline) reduce death or hospital readmission within 90 days?
• Study Type: Pragmatic, cluster-randomised, hospital-level crossover trial; open-label; administrative-data outcomes; seven hospitals in Ontario, Canada; two 12-week intervention periods per hospital.
• Population:
  • Setting: Whole-hospital (ED, wards, perioperative areas, and ICUs) within participating hospitals.
  • Inclusion: All hospital admissions during active trial periods (analysed as “index” admissions).
  • Key exclusions: Prior hospital admission within the previous 90 days (to define index admissions); age <1 month or missing birth date; not resident in Ontario; admissions during run-in/washout periods; data-quality exclusions (small numbers).
• Intervention:
  • Hospital-wide default use of lactated Ringer’s solution for intravenous crystalloid therapy during assigned periods.
  • Implementation via logistics and prescribing supports (e.g., stocking the assigned fluid to achieve an inventory target, and enabling default ordering), while allowing clinician override when clinically indicated.
  • Target operational separation included maintaining ≥80% of available crystalloid inventory as the assigned trial fluid.
• Comparison:
  • Hospital-wide default use of 0.9% normal saline for intravenous crystalloid therapy during assigned periods, with the same operational approach and allowance for clinician override.
• Blinding: Unblinded at clinician and site level (policy intervention); outcomes derived from routinely collected administrative data, reducing susceptibility of outcome ascertainment to performance/detection bias for objective endpoints.
• Statistics: Planned sample size of 12 hospitals to detect a 1.0 percentage point absolute reduction in the primary outcome (assumed 16% to 15%) with 80% power at a two-sided 5% significance level (cluster design assumptions included an intracluster correlation coefficient of 0.006 and cluster autocorrelation of 0.95); primary analysis by intention-to-treat using hospital-level aggregated outcomes and regression adjusted for calendar time, hospital size, and intervention period.
• Follow-Up Period: 90 days from the index admission for the primary composite and most secondary outcomes.

Key Results

This trial was stopped early. The planned 16-hospital trial was interrupted by the COVID-19 pandemic; seven hospitals completed both trial periods (total analysed index admissions: 22,017 in lactated Ringer’s periods and 21,609 in normal saline periods).

• Primary outcome was similar with lactated Ringer’s vs saline: 20.3% (SD 3.5) vs 21.4% (SD 3.3); adjusted relative difference 0.97 (95% CI 0.90 to 1.05); P=0.35.
• Secondary outcomes (90-day mortality, readmission, ED visits, new dialysis, discharge destination) were also similar; no serious adverse events were reported by participating hospitals.
• Pre-specified subgroups:
  • Age >80 years: risk difference −3.41 percentage points (95% CI −6.25 to −0.56) for the primary outcome.
  • Other subgroups were broadly consistent with the overall neutral estimate (e.g., ICU admission during index admission: risk difference −0.91; 95% CI −3.81 to 1.99).

Internal Validity

• Randomisation and allocation: Hospital-level randomisation used permuted blocks (block size 2); allocation was held centrally and disclosed to sites approximately 1 month before initiating the first period.
• Drop out / exclusions: Primary analysis included 43,626 index admissions (22,017 lactated Ringer’s; 21,609 saline); exclusions primarily reflected design definitions and data-linkage constraints rather than loss to follow-up.
• Post-randomisation exclusions: Admissions were excluded for being non-index (prior admission within 90 days; 1,211–1,934 per period), age <1 month or missing birth date (204–258 per period), non-Ontario residency (40–82 per period; values suppressed as ranges), and data-quality concerns (<6 per period).
• Performance / detection bias: Open-label delivery could influence co-interventions; however, outcomes were objectively captured via provincial administrative datasets, reducing subjectivity in ascertainment of death/readmission/dialysis.
• Protocol adherence: Separation was incomplete but material: assigned trial fluid comprised 78.2% of administered intravenous crystalloids during lactated Ringer’s periods vs 93.6% during saline periods.
• Baseline characteristics: Groups were well matched at scale (e.g., age 62.7±22.0 vs 62.2±22.0 years; ICU admission on day 1: 7.0% vs 7.1%; mechanical ventilation on day 1: 3.3% vs 3.4%).
• Heterogeneity: Only seven clusters completed both periods; crossover reduces between-hospital confounding, but inference is sensitive to cluster-level variability and period effects.
• Timing: The intervention was implemented at policy level at the start of each 12-week period; run-in/washout periods were used and excluded from analysis to mitigate carryover.
• Dose: Patient-level crystalloid volumes and indications were not the primary unit of analysis; primary contrast is the hospital-level default crystalloid environment rather than a fixed per-patient dose.
• Separation of the variable of interest: Assigned fluid use: 78.2% (lactated Ringer’s periods) vs 93.6% (saline periods) of administered intravenous crystalloids.
• Key delivery aspects: Intervention feasibility depended on inventory management and ordering defaults; clinician override preserved clinical discretion but attenuated exposure separation.
• Outcome assessment: Endpoints (death, readmission, ED visit, dialysis) are clinically important and largely objective within administrative data constraints.
• Statistical rigour: Intention-to-treat at cluster level with prespecified covariate adjustment; early stopping reduced the achieved sample (clusters) relative to design assumptions, widening uncertainty despite large patient counts.

Conclusion on Internal Validity: Moderate. The pragmatic cluster-crossover design and objective outcome capture support credible estimation of a policy-level effect, but early stopping (fewer clusters) and imperfect separation of crystalloid exposure materially limit precision and attenuate detectable effects.

External Validity

• Population representativeness: Broad, whole-hospital admissions (medical, surgical, maternity, paediatric, mental health) were included, making results relevant to formulary-level decisions beyond ICU-only populations.
• Key exclusions affecting generalisability: Exclusion of non-Ontario residents and infants <1 month may limit applicability to jurisdictions without comparable administrative linkage or neonatal populations.
• Health-system context: Conducted in Ontario hospitals with established procurement systems and administrative datasets; implementation may be harder where supply chains are constrained or where lactated Ringer’s availability is intermittent.
• Intervention scope: Findings apply to a “default crystalloid” policy with clinician override, not to mandated patient-level administration of a fixed fluid volume.

Conclusion on External Validity: Generally good for similar high-income hospital systems considering a default crystalloid policy, with reduced transferability to settings where supply, ordering infrastructure, or case-mix (e.g., neonates) substantially differ.

Strengths & Limitations

• Strengths: Pragmatic policy intervention aligned to real-world formulary decisions; large number of admissions across whole-hospital care pathways; cluster crossover to mitigate between-hospital confounding; objective, administrative-data outcomes with minimal loss to follow-up; prespecified subgroup analyses and covariate adjustment.
• Limitations: Stopped early with only seven clusters completing both periods, reducing power and precision; open-label delivery with potential co-intervention changes; incomplete exposure separation (78.2% vs 93.6% assigned fluid use); policy-level design cannot attribute effects to individual patient fluid receipt or volume; results may not extrapolate to other balanced solutions (e.g., Plasma-Lyte) or to settings with different procurement and prescribing systems.

Interpretation & Why It Matters

• Clinical practice

  In an unselected inpatient population, a hospital-wide lactated Ringer’s default policy did not materially change 90-day death/readmission compared with saline, suggesting that large outcome shifts from formulary substitution alone are unlikely.
• Policy and implementation

  The trial operationalises a real formulary question (stocking, order defaults, clinician override) and quantifies achievable separation in routine care, directly informing how “strong” a policy lever can be in practice.
• Research implications

  Large pragmatic trials with administrative follow-up can efficiently test ubiquitous interventions; future work should prioritise identifying high-risk subgroups and maximising exposure contrast where the expected effect is small.

Controversies & Subsequent Evidence

• How FLUID sits within the RCT landscape: FLUID extends prior balanced-crystalloid research to a whole-hospital policy setting; it complements ICU/ED pragmatic crossover trials and later large ICU trials that generally show small (or null) average effects, reinforcing that any benefit is likely modest and context-dependent.¹²³⁴
• Magnitude of plausible benefit: Contemporary evidence syntheses (including systematic review/meta-analysis and individual patient data meta-analysis) support, at most, small absolute effects of balanced crystalloids over saline in broad critically ill populations; FLUID’s early termination means its confidence intervals remain compatible with similarly small benefit or harm in an unselected inpatient cohort.⁵⁶
• Methodological contribution (policy trial design): FLUID’s published protocol and analysis plan, and the earlier pilot protocol, formalise a reproducible approach to hospital-wide crystalloid substitution trials (inventory targets, crossover structure, and administrative endpoint ascertainment).⁷⁸
• Guideline context: Major critical care guidelines generally favour balanced crystalloids over saline for resuscitation in many contexts; these recommendations were largely developed before FLUID’s publication and therefore FLUID primarily strengthens whole-hospital generalisability rather than changing direction of guidance.⁹¹⁰

Summary

• Pragmatic hospital-wide, cluster-crossover trial comparing default lactated Ringer’s vs default normal saline across all admissions in seven Ontario hospitals.
• Stopped early due to the COVID-19 pandemic (planned 16 hospitals; seven completed both periods), yielding 43,626 index admissions in the primary analysis.
• No statistically significant difference in the primary composite of death or readmission at 90 days: 20.3% (SD 3.5) vs 21.4% (SD 3.3); adjusted relative difference 0.97 (95% CI 0.90 to 1.05); P=0.35.
• Secondary outcomes (including 90-day mortality and new dialysis) were similar; no serious adverse events were reported.
• Exposure separation was incomplete but clinically meaningful at policy level (assigned fluid 78.2% of crystalloids in lactated Ringer’s periods vs 93.6% in saline periods).

Further Reading

Other Trials

• 2018Semler MW, Self WH, Wanderer JP, et al. Balanced crystalloids versus saline in critically ill adults. N Engl J Med. 2018;378(9):829-839.
• 2018Self WH, Semler MW, Wanderer JP, et al. Balanced crystalloids versus saline in noncritically ill adults. N Engl J Med. 2018;378(9):819-828.
• 2022Finfer S, Micallef S, Hammond N, et al. Balanced multielectrolyte solution versus saline in critically ill adults. N Engl J Med. 2022;386(9):815-826.
• 2021Zampieri FG, Machado FR, Biondi RS, et al. Effect of intravenous fluid treatment with a balanced solution vs saline solution on mortality in critically ill patients: the BaSICS randomized clinical trial. JAMA. 2021;326(9):1-12.
• 2015Young P, Bailey M, Beasley R, et al. Effect of a buffered crystalloid solution vs saline solution on acute kidney injury among patients in the intensive care unit: the SPLIT Randomized Clinical Trial. JAMA. 2015;314(16):1701-1710.

Systematic Review & Meta Analysis

• 2022Hammond NE, Zampieri FG, Di Tanna GL, et al. Balanced crystalloids versus saline in critically ill adults—a systematic review with meta-analysis. NEJM Evid. 2022;1(2):1-12.
• 2024Zampieri FG, Cavalcanti AB, Di Tanna GL, et al. Balanced crystalloids versus saline in critically ill patients: an individual patient data meta-analysis. Lancet Respir Med. 2024;12:237-246.
• 2018Kawano-Dourado L, Zampieri FG, Azevedo LCP, et al. Low- versus high-chloride content intravenous solutions for critically ill and perioperative adult patients: a systematic review and meta-analysis. Anesth Analg. 2018;126(2):513-521.
• 2012Burdett E, Dushianthan A, Bennett-Guerrero E, et al. Perioperative buffered versus non-buffered fluid administration for surgery in adults. Cochrane Database Syst Rev. 2012;(12):CD004089.
• 2014Myburgh JA, Mythen MG. Resuscitation fluids. N Engl J Med. 2013;369:1243-1251.

Observational Studies

• 2012Shaw AD, Bagshaw SM, Goldstein SL, et al. Major complications, mortality, and resource utilisation after open abdominal surgery: 0.9% saline compared to Plasmalyte. Ann Surg. 2012;255(5):821-829.
• 2013McCluskey SA, Karkouti K, Wijeysundera D, et al. Hyperchloremia after noncardiac surgery is independently associated with increased morbidity and mortality: a propensity-matched cohort study. Anesth Analg. 2013;117(2):412-421.
• 2012Yunos NM, Bellomo R, Hegarty C, Story D, Ho L, Bailey M. Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012;308(15):1566-1572.
• 2011Yunos NM, Kim IB, Bellomo R, et al. The biochemical effects of restricting chloride-rich fluids in intensive care. Crit Care Med. 2011;39(11):2419-2424.
• 2012Chowdhury AH, Cox EF, Francis ST, Lobo DN. A randomised, controlled, double-blind crossover study on the effects of 2-L infusions of 0.9% saline and Plasma-Lyte 148 on renal blood flow velocity and renal cortical tissue perfusion in healthy volunteers. Ann Surg. 2012;256:18-24.

Guidelines

• 2024Arabi YM, et al. European Society of Intensive Care Medicine clinical practice guidelines on fluid therapy in adult critically ill patients: part 1: resuscitation fluids. Intensive Care Med. 2024.
• 2025European Society of Intensive Care Medicine. Clinical practice guidelines on fluid therapy in adult critically ill patients: part 2: volume for resuscitation and replacement fluids. Intensive Care Med. 2025.
• 2025European Society of Intensive Care Medicine. Clinical practice guidelines on fluid therapy in adult critically ill patients: part 3: fluid removal. Intensive Care Med. 2025.
• 2021Evans L, Rhodes A, Alhazzani W, et al. Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47:1181-1247.
• 2021Evans L, Rhodes A, Alhazzani W, et al. Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2021. Crit Care Med. 2021;49:e1063-e1143.

Notes

• ESICM guideline evidence syntheses were completed with literature searches up to February 2025, so they may not incorporate FLUID’s 2025 publication.

Overall Takeaway

FLUID is a landmark implementation-focused RCT because it tests crystalloid choice as a hospital policy rather than a patient-level prescription, across the full spectrum of inpatient care. Its neutral primary result suggests that hospital-wide substitution of lactated Ringer’s for saline is unlikely to produce large population-level reductions in death/readmission, while still leaving room for modest effects (or subgroup-specific benefits) that would require larger numbers of clusters and stronger exposure separation to detect.

Bibliography

• 1Semler MW, Self WH, Wanderer JP, et al. Balanced crystalloids versus saline in critically ill adults. N Engl J Med. 2018;378(9):829-839.
• 2Self WH, Semler MW, Wanderer JP, et al. Balanced crystalloids versus saline in noncritically ill adults. N Engl J Med. 2018;378(9):819-828.
• 3Finfer S, Micallef S, Hammond N, et al. Balanced multielectrolyte solution versus saline in critically ill adults. N Engl J Med. 2022;386(9):815-826.
• 4Zampieri FG, Machado FR, Biondi RS, et al. Effect of intravenous fluid treatment with a balanced solution vs saline solution on mortality in critically ill patients: the BaSICS randomized clinical trial. JAMA. 2021;326(9):1-12.
• 5Hammond NE, Zampieri FG, Di Tanna GL, et al. Balanced crystalloids versus saline in critically ill adults—a systematic review with meta-analysis. NEJM Evid. 2022;1(2):1-12.
• 6Zampieri FG, Cavalcanti AB, Di Tanna GL, et al. Balanced crystalloids versus saline in critically ill patients: an individual patient data meta-analysis. Lancet Respir Med. 2024;12:237-246.
• 7Shaw JF, Chasse M, Adhikari NKJ, Lamontagne F, McIntyre L, McArdle T, et al. Protocol and Statistical Analysis Plan of FLUID Trial: A Pragmatic Cluster Crossover Trial of Lactated Ringer’s Solution Versus Saline in Hospitalised Patients. JMIR Res Protoc. 2023 Oct 6;12:e51783.
• 8McIntyre LA, Chasse M, English SW, et al. The FLUID pilot trial: a protocol for a pilot cluster randomized crossover trial of the hospital-wide application of lactated ringer's solution versus normal saline. BMJ Open. 2018;8:e022780.
• 9Arabi YM, et al. European Society of Intensive Care Medicine clinical practice guidelines on fluid therapy in adult critically ill patients: part 1: resuscitation fluids. Intensive Care Med. 2024.
• 10Evans L, Rhodes A, Alhazzani W, et al. Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47:1181-1247.