1. Publication
- Title: A Crossover Trial of Hospital-Wide Lactated Ringer's Solution versus Normal Saline (FLUID Trial)
- Acronym: FLUID
- Year & Journal: New England Journal of Medicine, published August 14, 2025 (2025;393(7):660-670)
- Citation: McIntyre L, Fergusson D, McArdle T, et al; Canadian Critical Care Trials Group. A Crossover Trial of Hospital-Wide Lactated Ringer's Solution versus Normal Saline. N Engl J Med. 2025;393(7):660-670. doi:10.1056/NEJMoa2416761
2. Context & Rationale
Background: Balanced crystalloids (e.g., lactated Ringer's) versus saline has produced mixed results across settings: a systematic review of 13 ICU trials (35,884 patients) found no significant mortality difference, while a Bayesian meta-analysis of ~35,000 ICU patients suggested a high probability of lower mortality with balanced fluids. Prior large trials (BASICS, PLUS) in ICU populations found no significant 90-day mortality difference. Whether a hospital-wide default fluid policy (not just ICU-specific) matters for broader inpatient outcomes was untested.
Research Question/Hypothesis: Across an entire hospital (not ICU-restricted), does using lactated Ringer's solution as the default IV fluid reduce the composite of death or 90-day readmission compared with normal saline?
Why This Matters: Tests whether a simple, hospital-wide policy change (switching the default crystalloid) meaningfully affects broad inpatient outcomes — a pragmatic, system-level question distinct from prior ICU-specific balanced-fluid trials.
3. Design & Methods
- Study Type: Open-label, two-period, two-sequence, cross-sectional, cluster-randomized, crossover trial
- Setting & Centers: Hospitals in Ontario, Canada; conducted 2016-2020
- Population:
- Inclusion: All hospitalized patients receiving IV crystalloid at participating hospitals during each 12-week study period
- Design: Hospitals used one fluid hospital-wide for 12 weeks, followed by washout, then switched to the other fluid for 12 weeks
- Intervention: Lactated Ringer's solution as default hospital-wide IV fluid
- Comparator: Normal saline (0.9% NaCl) as default hospital-wide IV fluid
- Randomization: Cluster (hospital)-level, cross-sectional, crossover design; 7 hospitals completed both 12-week periods before the trial was interrupted by the COVID-19 pandemic
- Blinding: Open-label
- Statistical Power & Follow-Up: Primary outcome: composite of death or hospital readmission within 90 days after index admission, obtained from health administrative databases. Secondary: individual components, length of stay, dialysis initiation, ED visits, discharge destination. 43,626 eligible patients with primary-outcome data (22,017 LR, 21,609 NS).
4. Key Results
43,626 eligible patients across 7 hospitals.
Outcome | Lactated Ringer's | Normal Saline | Effect Size | 95% CI | p-value | Notes |
Death or 90-day readmission (primary) | 20.3% ±3.5 (hospital-level mean±SD) | 21.4% ±3.3 | Adjusted diff −0.53 pp | −1.85 to 0.79 | 0.35 | No significant difference |
All secondary outcomes | Consistent with primary | Consistent with primary | Not reported | Not reported | — | No significant differences; no serious adverse events attributable to either fluid |
5. Internal Validity Assessment
- Randomization & Allocation: Cluster (hospital)-level randomization with cross-sectional, crossover design and washout periods — each hospital serves as its own control, controlling for hospital-level confounders.
- Protocol Adherence & Separation: Hospital-wide default-fluid policy ensures strong real-world adherence (the intervention is embedded in the ordering system, not dependent on individual clinician choice per patient).
- Blinding & Detection Bias: Open-label (inherent to a system-level fluid-policy intervention); outcomes obtained from objective health administrative databases, limiting detection bias.
- Missing Data & Sensitivity Analyses: Trial interrupted by COVID-19 pandemic after 7 hospitals completed both periods (fewer than originally planned) — a real-world limitation affecting final sample size, though still yielding a very large (43,626-patient) analysis.
- Overall Internal Validity Conclusion: Strong — despite pandemic-related early interruption, the trial's cluster-crossover design (each hospital as its own control) and very large sample size (43,626 patients) provide a robust, precisely estimated null result, with a narrow confidence interval around a small effect size.
6. External Validity Assessment
- Population Representativeness: All hospitalized patients receiving IV crystalloid at 7 Ontario hospitals — a genuinely broad, hospital-wide (not ICU-restricted) population, distinct from and complementary to prior ICU-specific balanced-fluid trials.
- Practice Context: Directly tests a hospital-wide default-fluid policy, the exact intervention a hospital pharmacy/formulary committee would consider implementing.
- Overall External Validity Conclusion: Excellent for hospital-wide fluid policy decisions — this is precisely the pragmatic, system-level question hospital administrators and pharmacy committees need answered, distinct from narrower ICU-specific trials.
7. Strengths & Limitations
Strengths:
- Very large sample (43,626 patients) with a sophisticated cluster-crossover design controlling for hospital-level confounding
- Directly answers a genuinely different, system-level question (hospital-wide default policy) than prior ICU-specific balanced-fluid trials
- Precise, well-powered null result
- No serious adverse events attributable to either fluid
Limitations:
- Open-label design (though administrative-database outcome ascertainment limits impact)
- Trial interrupted by COVID-19 pandemic, resulting in fewer completed hospital-periods than originally planned
- Single-province (Ontario, Canada) trial
- Broad hospital-wide population dilutes any potential effect specific to higher-severity subgroups (e.g., septic shock) where prior trials (SMART) suggested balanced-fluid benefit
8. Interpretation & Practice Impact
- Clinical Implications: A hospital-wide policy of using lactated Ringer's rather than normal saline as the default IV fluid does not meaningfully change death or 90-day readmission across the general hospitalized population — hospitals need not prioritize a system-wide fluid-policy change purely for this outcome.
- Mechanistic Coherence: Consistent with prior large ICU-specific balanced-fluid trials (BASICS, PLUS) that also found no significant 90-day mortality difference, though this trial extends the null finding to the broader, less severely ill hospital-wide population.
- Systems-Level Takeaway: Hospital-wide institution of a default IV crystalloid solution may not meaningfully change patient outcomes at the population level, though this does not preclude potential benefit in specific higher-risk subgroups (e.g., septic shock, as suggested by SMART) that a hospital-wide, unselected trial cannot isolate.
9. Controversies & Subsequent Evidence
- Editorial Commentary/Debates: Independent commentary (The Hospitalist) explicitly frames the bottom line as "hospital-wide institution of a default IV crystalloid solution may not change meaningful patient outcomes," while noting this contrasts with more favorable signals for balanced fluids in higher-severity, ICU-specific populations (SMART trial, Bayesian meta-analyses).
- Guideline Integration: Complements rather than contradicts prior ICU-specific balanced-fluid evidence — the key nuance is that broad, unselected hospital-wide populations may show no net effect even if higher-severity subgroups benefit, since dilution across a large low-risk population can mask a smaller, real effect in the sickest patients.
10. Summary & Executive Takeaway
Summary: This cluster-randomized, crossover trial across 7 Ontario hospitals compared hospital-wide default use of lactated Ringer's versus normal saline in 43,626 patients. The composite of death or 90-day readmission was not significantly different (20.3% vs 21.4%; adjusted difference −0.53 percentage points, 95% CI −1.85 to 0.79, P=0.35), with no differences in any secondary outcome.
Overall Takeaway: A hospital-wide policy of using lactated Ringer's rather than normal saline as the default IV fluid does not meaningfully improve broad inpatient outcomes — hospital-wide fluid policy decisions may be reasonably guided by cost, availability, and local preference, while more severely ill subgroups (e.g., septic shock) may still warrant individualized balanced-fluid consideration based on prior ICU-specific trial evidence.
11. Bibliography
- Brown RM, Wang L, Coston TD, et al. Balanced Crystalloids versus Saline in Critically Ill Adults (SMART). N Engl J Med. 2018;378:829-839.
- Zampieri FG, et al. Effect of Intravenous Fluid Treatment with a Balanced Solution vs 0.9% Saline Solution on Mortality in Critically Ill Patients (BASICS). JAMA. 2021;326(9):818-829.
- Finfer S, et al. Balanced Multielectrolyte Solution versus Saline in Critically Ill Adults (PLUS). N Engl J Med. 2022;386:815-826.
- Shaw JF, Ouyang Y, Fergusson DA, et al. The FLUID Trial protocol. JMIR Res Protoc. 2023.