FACTT

Published

Wiedemann HP, et al. "Comparison of two fluid-management strategies in acute lung injury". The New England Journal of Medicine. 2006. 354(24):2564-75.
PubMed • Full text • PDF

Clinical Question

In patients with ALI/ARDS that are intubated and receiving positive pressure ventilation, how does the conservative compare to the liberal fluid management strategy in reducing mortality?

Bottom Line

Among patients with ALI/ARDS, a conservative fluid management strategy targeting a CVP <4 mmHg improves lung function, decreases ventilator days, and reduces ICU days compared to a liberal strategy.

Major Points

Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) are associated with a high mortality -- up to 46% at 60 days for ARDS. The only major improvement in mortality had been the ARDSNet (2000) low tidal volume protocol. Pulmonary edema plays a significant role in the ALI/ARDS disease process, reducing the degree of pulmonary edema through a conservative fluid strategy may benefit in reducing mortality.

The 2006 Fluids and Catheters Treatment Trial (FACTT) sought to define the optimal fluid management strategy in ALI/ARDS. It randomized 1,000 patients with ALI/ARDS to liberal (CVP 10-14) or conservative (CVP <4) fluid management strategies. At 60 days, there was no difference between the two groups in mortality. However, the conservative therapy was associated with a decreased days on a ventilator and time in the ICU.

Of note, the trial also compared use of central venous catheters (CVC) and pulmonary artery catheters (PAC) in monitoring central pressure. There was no difference in mortality at 60 days between the two groups. Routine PAC use has fallen out of favor because of the lack of mortality benefit with increased rates of complications in acute HF management (ESCAPE; 2005) and ICU patients (PAC-MAN; 2005).

Guidelines

Surviving Sepsis Campaign (2012)

• Conservative fluid strategy for patients with known sepsis-induced ARDS without tissue hypoperfusion (grade 1C)

Design

• Multicenter, two-by-two factorial, randomized controlled trial
• N=1,000
  • Conservative (n=503)
  • Liberal (n=497)
• Setting: 20 centers in N. America
• Enrollment: 2000-2005
• Follow-up: 60 days
• Analysis: Intention-to-treat
• Primary outcomes:
  • All-cause mortality at 60 days
  • Dialysis at day 60

Population

Inclusion Criteria

• Intubated patients receiving positive-pressure ventilation
• All patients had ALI (defined as PaO₂:FiO₂ ratio <300, bilateral infiltrates on CXR, and no evidence of LA hypertension)
• CVC or plan to insert CVC

Exclusion Criteria

• PA catheter after onset of ALI
• ALI >48h
• Inability to obtain consent
• Chronic conditions that independently influence survival, impair weaning, or compromise compliance
• Irreversible conditions with 6 month mortality >50%

Baseline Characteristics

From the conservative fluid management group.

• Demographics: Age 50 years, male sex 52%, white race 65%, black race 20%, Hispanic 12%
• Primary lung injury: Pneumonia 46%, sepsis 22%, aspiration 16%, trauma 8%, multiple transfusions 1%
• PMH: DM 18%, HIV +/- AIDS 7%, cirrhosis 3%, solid tumors 1%, leukemia 3%, lymphoma 2%, immunosuppression 9%
• Baseline health data: APACHE III score 93.1, MAP 77.1 mmHg, CVP 11.9 mmHg, APOP 15.6 mmHG, CI 4.2 L/min/m², SVO₂ 69%, shock 33%, vasopressors 31%
• Baseline respiratory data: Vt 7.4 mL/kg of PBW, plateau pressure 26.2 cm H₂O, PaO₂:FiO₂157, PEEP 9.4 cm H₂O, pH 7.36
• Baseline labs: BUN 23.2 mg/dL, creatinine 1.24 mg/dL, bicarbonate 22.5 mmol/L, Hgb 10.4 g/dL, glucose 138 mg/dL
• Fluid balance before randomization: +2655 mL
• Medical ICU: 66%

Interventions

• All patients were on ARDS Network (ARDSNet) protocol (low Vt).
• Randomized to conservative vs. liberal fluid management strategy, and randomized to CVC or PAC
  • Suspension of diuretic administration if in shock
• Received assigned catheter within four hours of randomization. PAC could be replaced with CVC if hemodynamically stable.

Outcomes

Comparisons are conservative vs. liberal fluid management strategy except where stated.

Primary Outcomes

All-cause mortality at 60 days

25.5% vs. 28.4% (P=0.30)

By catheter type: No difference (data not presented, P=0.24)

Dialysis at day 60

10% vs. 14% (P=0.06)

Secondary Outcomes

Ventilator-free days

14.6% vs. 12.1% (P<0.001)

Mechanical ventilation duration

Mean: 10.37 vs. 13.59 days (P<0.001)

Median: 6 vs. 9 days (P<0.001)

ICU-free days

Within first week: 0.9% vs. 0.6% (P<0.001)

Within first month: 13.4% vs. 11.2% (P<0.001)

CV failure-free days within first week

3.9% vs. 4.2% (P=0.04)

CNS failure-free days

Within first week: 3.4% vs. 2.9% (P=0.02)

Within first month: 18.8% vs. 17.2% (P=0.03)

Additional Analyses

Daily furosemide dose

Day 1: 149 vs. 74 mg (P<0.001)

Day 7: 137 vs. 87 mg (P<0.001)

Fluid balance

Cumulative -136 vs. +6992 mL (P<0.001)

Day 1: +1187 vs. +2529 mL (P<0.001)

Day 2: -376 vs. +1643 mL (P<0.001)

Day 3: -408 vs. +936 mL (P<0.001)

Day 4: -165 vs. +564 mL (P<0.001)

Day 5: -226 vs. +483 mL (P<0.001)

Day 6: -145 vs. +508 mL (P<0.001)

Day 7: +130 v.s +459 mL (P=0.04)

Subgroup Analysis

There was no difference for the primary outcome or ventilator-free days for presence of baseline shock. Hispanic participants had a higher rate of mortality than white participants (HR 1.58; 95% CI 1.08-2.31; P not given).

Criticisms

• The conservative group was treated with a diuretic regimen that resulted in a net-even (rather than net-negative) fluid status

Funding

Funding provided by the National Heart, Lung, and Blood Institute.