Publication

• Title: Comparison of Two Fluid-Management Strategies in Acute Lung Injury
• Acronym: FACTT (Fluid and Catheter Treatment Trial)
• Year: 2006
• Journal published in: New England Journal of Medicine
• Citation: Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, deBoisblanc B, Connors AF Jr, Hite RD, Harabin AL; National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med. 2006;354(24):2564-2575.

Context & Rationale

• Background

  • Acute lung injury (ALI)/ARDS is characterised by increased alveolar-capillary permeability; hydrostatic pressure and total fluid balance can modulate oedema burden and gas exchange.
  • Clinical practice historically oscillated between “haemodynamic support” (liberal fluids) and “lung protection” (fluid restriction/diuresis), with major concern that fluid restriction could precipitate shock and renal injury.
  • Prior evidence was dominated by physiology and observational associations; a large, protocolised randomised trial was lacking.
• Research Question/Hypothesis

  • In adults with ALI, does a conservative fluid-management strategy (targeting lower filling pressures) reduce death before discharge home by day 60 and improve clinical recovery (e.g., ventilator-free days), compared with a liberal strategy?
  • Does the effect of fluid strategy depend on the haemodynamic catheter used (pulmonary-artery catheter vs central venous catheter) within the parent factorial trial?
• Why This Matters

  • Fluid strategy is a ubiquitous, modifiable ICU intervention with plausible, competing risks (lung oedema vs organ hypoperfusion).
  • Endpoints such as ventilator-free and ICU-free days translate directly into patient-centred recovery and ICU resource utilisation.
  • FACTT provided foundational evidence for “late conservative / deresuscitative” care in ALI/ARDS once immediate shock is addressed.

Design & Methods

• Research Question: Among patients with ALI, does a conservative fluid strategy (vs liberal) improve 60-day clinical outcomes (primary: death before discharge home by day 60; key secondary: ventilator-free and ICU-free days), without excess shock or renal replacement therapy?
• Study Type: Randomised, multicentre (20 North American centres), investigator-initiated NIH/ARDSNet trial; factorial design with simultaneous randomisation to haemodynamic catheter (pulmonary-artery vs central venous) and to fluid-management strategy; open-label protocolised intervention; enrolment June 2000 to October 2005.
• Population:
  • Setting: Mechanically ventilated ICU patients with ALI/ARDS managed within ARDSNet protocols.
  • Key inclusion (ALI definition): bilateral infiltrates consistent with pulmonary oedema; PaO₂:FIO₂ ≤300; no evidence of left atrial hypertension (or, if measured, pulmonary capillary wedge pressure ≤18 mm Hg); acute onset.
  • Timing constraints: randomisation permitted only within 48 hours after ALI criteria first met; catheter to be inserted within 4 hours of randomisation; first fluid-management protocol instruction within 2 hours after catheter placement.
  • Key exclusions (high-level): pre-existing pulmonary-artery catheter; severe chronic lung disease; severe chronic liver disease (Child–Pugh 10–15); chronic dialysis dependence; severe chronic heart failure; anticipated death within 6 months; inability/unwillingness to pursue full aggressive care; other protocol-defined exclusions.
• Intervention:
  • Conservative fluid-management strategy (7-day protocol): target lower intravascular filling pressures (CVP <4 mm Hg if central venous catheter; PAOP <8 mm Hg if pulmonary-artery catheter).
  • Operationalisation: protocol-directed 4-hourly reassessments using a decision support algorithm incorporating (a) intravascular pressure, (b) evidence of shock, and (c) evidence of ineffective circulation.
  • Typical actions when haemodynamically suitable: fluid restriction and diuresis (e.g., furosemide), aiming for a near-neutral/negative balance while avoiding hypoperfusion.
• Comparison:
  • Liberal fluid-management strategy (7-day protocol): target higher intravascular filling pressures (CVP 10–14 mm Hg if central venous catheter; PAOP 14–18 mm Hg if pulmonary-artery catheter).
  • Operationalisation: the same structured 4-hourly reassessment framework, with protocol actions favouring fluid administration and reduced diuresis to maintain higher filling pressures, while addressing shock/ineffective circulation when present.
• Blinding: Unblinded (treating teams applied haemodynamic targets and fluid/diuretic actions); primary endpoint (death before discharge home by day 60) is objective; secondary endpoints (ventilator-/ICU-free days) are partially care-process dependent but were assessed within an ARDSNet protocol environment.
• Statistics: Sample size planned at 1000 patients to provide 90% power (two-sided α=0.05) to detect a 10 percentage-point absolute reduction in the primary endpoint (from 31% to 21%); primary analyses performed according to intention-to-treat principles (with one post-randomisation withdrawal before intervention exposure excluded from analysis as reported).
• Follow-Up Period: Outcomes assessed through day 60 for mortality/discharge home status and dialysis; ventilator-free/ICU-free days assessed to day 28; protocol intervention applied for 7 days after randomisation.

Key Results

This trial was not stopped early. Enrolment reached the planned sample size (1000 randomised; 503 conservative and 497 liberal analysed).

• Mortality was similar (25.5% conservative vs 28.4% liberal by day 60; P=0.30) despite substantial protocol separation in cumulative fluid balance at day 7 (−136 ± 491 mL vs 6992 ± 502 mL; P<0.001).
• Conservative strategy improved recovery surrogates: ventilator-free days (14.6 ± 0.5 vs 12.1 ± 0.5; P<0.001) and ICU-free days (13.4 ± 0.4 vs 11.2 ± 0.4; P<0.001).
• Renal replacement therapy was not increased (dialysis to day 60: 10% vs 14%; P=0.06), but electrolyte imbalance events were more frequent with conservative management (42% vs 19%; P=0.001).

Internal Validity

• Randomisation and allocation: Centralised computer-generated randomisation with permuted blocks; factorial design; allocation concealment supported by automated assignment.
• Dropout/exclusions: 1001 randomised; 1 patient in the liberal group withdrew consent before receiving the intervention and was excluded from analysis; analysed cohorts were 503 (conservative) and 497 (liberal); mortality follow-up to day 60 was otherwise complete.
• Performance/detection bias: Open-label delivery; primary outcome objective; secondary outcomes (ventilator-/ICU-free days) potentially influenced by care processes, mitigated by protocolised ARDSNet management environment.
• Protocol adherence and separation: Large between-group separation achieved in fluid balance: cumulative day-7 balance −136 ± 491 mL (conservative) vs 6992 ± 502 mL (liberal).
• Separation of the variable of interest (examples with published daily values): Day-1 fluid balance 1186.7 ± 151.01 mL (conservative) vs 2529.5 ± 148.99 mL (liberal); day-2 −376.1 ± 161.08 mL vs 1642.9 ± 151.71 mL; day-3 −1171.0 ± 180.80 mL vs 1495.7 ± 143.40 mL.
• Dose/intensity of protocol actions: Higher furosemide dosing in conservative group on each study day (e.g., day 1: 148.94 ± 8.52 mg/24h vs 74.27 ± 7.48 mg/24h).
• Baseline comparability: Groups were well balanced across major prognostic factors (e.g., APACHE III score 93.1 ± 1.4 vs 95.2 ± 1.4; shock criteria at baseline 33% vs 36%; PaO₂:FIO₂ 179.8 ± 4.1 vs 176.7 ± 4.2).
• Heterogeneity and interaction: No significant interaction between the catheter intervention and fluid strategy; no significant interaction between baseline shock status and fluid strategy with respect to primary/major secondary outcomes (as reported).
• Timing: Randomisation occurred after patients met ALI criteria (within 48 hours by protocol), and fluid strategy began after haemodynamic catheter placement; prerandomisation fluid balance was already positive (e.g., 2655 ± 156 mL vs 2875 ± 166 mL).
• Outcome assessment: Mortality/discharge home status through day 60; ventilator-/ICU-free days and organ-failure–free days through day 28; dialysis through day 60.
• Statistical rigour: Planned sample size achieved; primary analysis aligned with prespecified power assumptions and intention-to-treat framework.

Conclusion on Internal Validity: Strong overall: randomisation and follow-up were robust, protocol separation in fluid balance was large and sustained, and the primary outcome was objective, although open-label delivery could influence process-dependent secondary endpoints.

External Validity

• Population representativeness: Adults with ALI in academic/tertiary ICUs within a North American clinical trials network; exclusion criteria removed patients with advanced chronic organ disease (e.g., severe chronic respiratory disease, chronic dialysis, severe chronic liver disease) and those not committed to full support.
• Practice context: The intervention relied on frequent reassessment and haemodynamic targets (CVP or PAOP depending on catheter assignment), which may not reflect contemporary practice where pulmonary-artery catheters are uncommon and dynamic assessments are favoured.
• Applicability across phases of illness: Findings most directly apply after initial stabilisation/resuscitation in established ALI/ARDS, rather than the earliest shock-resuscitation phase.
• Health-system transferability: Protocolised strategies are feasible in well-resourced ICUs; implementation in resource-limited settings may be constrained by staffing, monitoring, and laboratory availability.

Conclusion on External Validity: Moderately strong for protocol-capable ICUs managing established ALI/ARDS, but less directly transferable to settings without structured haemodynamic monitoring workflows or to the earliest resuscitation phase of shock.

Strengths & Limitations

• Strengths: Large multicentre randomised trial; protocolised and clearly differentiated fluid strategies; clinically meaningful secondary outcomes (ventilator-free and ICU-free days); high completeness of follow-up; factorial design permitted assessment of catheter–strategy interaction.
• Limitations: Open-label intervention; reliance on filling-pressure targets (CVP/PAOP) and intensive protocol infrastructure; prerandomisation fluid exposure substantial (positive balance at baseline); mortality endpoint potentially insensitive to smaller effect sizes; interpretation of renal function complicated by haemoconcentration during conservative management.

Interpretation & Why It Matters

• Clinical signal

  In established ALI/ARDS, a protocolised conservative strategy achieved near-neutral fluid balance (−136 ± 491 mL at day 7) and improved ventilator-free and ICU-free days without a statistically significant increase in dialysis use.
• How practice changed

  FACTT underpins the common ICU paradigm of prioritising haemodynamic stabilisation early, then transitioning to conservative fluid management/deresuscitation during the pulmonary oedema phase to facilitate liberation from ventilation.
• Risk trade-off

  Conservative management increased metabolic/electrolyte events (42% vs 19%), reinforcing the need for surveillance (electrolytes, acid–base) during aggressive diuresis and fluid restriction.

Controversies & Subsequent Evidence

• Phase-of-illness framing: The trial tested a protocol applied after ALI was established (often following substantial prerandomisation fluids), and contemporaneous commentary emphasised the distinction between early resuscitation and later conservative/deresuscitative phases rather than a single “always restrictive” rule.¹
• Renal signal and creatinine interpretation: Conservative strategy produced haemoconcentration (higher creatinine/BUN over time as reported), raising debate about whether “renal safety” is best judged by dialysis alone versus creatinine-based AKI definitions that are sensitive to fluid balance; a later analysis demonstrated that adjusting creatinine for fluid accumulation materially alters AKI classification and its relationship with outcomes in ALI cohorts.²
• Implementability outside trials: The original FACTT algorithm required frequent reassessment and catheter-derived targets; a simplified conservative protocol has been evaluated to improve feasibility while maintaining a conservative approach in ARDS care pathways.³
• Effect size and power: FACTT was powered for a 10% absolute mortality reduction (31% to 21%); the observed mortality difference between strategies was smaller, so absence of statistical significance does not exclude more modest mortality effects.
• Modern synthesis and guidance: Subsequent reviews and guidance documents commonly incorporate FACTT by supporting conservative fluid management after haemodynamic stabilisation in ARDS, while emphasising careful monitoring for electrolyte/acid–base derangements and renal effects (where definitions may be fluid-sensitive).

Summary

• In 1000 patients with ALI, conservative fluid management did not significantly reduce 60-day mortality compared with liberal management (25.5% vs 28.4%; P=0.30).
• Conservative strategy produced major protocol separation in fluid balance at day 7 (−136 ± 491 mL vs 6992 ± 502 mL; P<0.001).
• Conservative strategy increased ventilator-free days (14.6 ± 0.5 vs 12.1 ± 0.5; P<0.001) and ICU-free days (13.4 ± 0.4 vs 11.2 ± 0.4; P<0.001).
• Dialysis use to day 60 was not higher with conservative management (10% vs 14%; P=0.06), but electrolyte imbalance events were more frequent (42% vs 19%; P=0.001).
• FACTT is a cornerstone trial supporting conservative/deresuscitative fluid management in established ARDS once shock and ineffective circulation are addressed.

Overall Takeaway

FACTT demonstrated that, in established ALI/ARDS managed in protocolised ICUs, targeting lower filling pressures and achieving near-neutral fluid balance improves ventilator-free and ICU-free days without a statistically significant mortality benefit. The trial reshaped modern supportive care by legitimising conservative/deresuscitative fluid management once shock and ineffective circulation are addressed, while highlighting the need for vigilance regarding electrolyte and renal physiology.

Bibliography

• Rivers EP. Fluid-management strategies in acute lung injury — liberal, conservative, or both? N Engl J Med. 2006;354:2598-2600.
• Liu KD, Thompson BT, Ancukiewicz M, et al. Acute kidney injury in patients with acute lung injury: impact of fluid accumulation on classification of acute kidney injury and associated outcomes. Crit Care Med. 2011;39(12):2665-2671.
• Grissom CK, Hirshberg EL, Dickerson JB, Brown SM, Lanspa MJ, Liu KD, et al. Fluid management with a simplified conservative protocol for the acute respiratory distress syndrome. Crit Care Med. 2015;43(2):288-295.