FLORALI

In Review

Frat JP, et al. "High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure". The New England Journal of Medicine. 2015. 372(23):2185-2196.
PubMed • Full text • PDF • ClinicalTrials.gov

Clinical Question

In patients with acute hypoxemic respiratory failure without hypercapnia, does high-flow oxygen therapy decrease intubation and mortality rates compared to noninvasive ventilation or standard oxygen therapy?

Bottom Line

ICU and 90-day mortality were significantly reduced in patients with acute hypoxemic respiratory failure without hypercapnia given high-flow oxygen therapy compared to noninvasive ventilation or standard oxygen therapy.

Major Points

There are multiple methods of delivering oxygen that aim to minimize intubation rate and mortality, such as nasal cannula, non-rebreather mask, and noninvasive positive-pressure ventilation (NIPPV) such as BiPAP and CPAP. There is high quality data that NIPPV reduces intubation rate and mortality in COPD exacerbations and cardiogenic pulmonary edema, however its benefits in other forms of acute hypoxemic respiratory failure have yet to be clearly demonstrated.

The FLORALI trial (2015) demonstrated the effectiveness of a new method of oxygen delivery called high-flow nasal cannula in a well-defined subset of patients in acute respiratory failure: those with hypoxemia and without hypercapnia. In this randomized controlled trial, high-flow oxygen therapy had a significant improvement in 90-day mortality compared to either noninvasive ventilation or standard oxygen therapy (12% vs. 28% vs. 23%). While there was no significant reduction in intubation rate overall, patients with more severe hypoxemia, PaO2:FiO2 ≤ 200 mmHg, did experience a significant reduction in intubation rates with high-flow oxygen therapy, which may have driven this mortality benefit.

In contrast with FLORALI, the 2018 HIGH trial did not find a mortality benefit for HFNC versus standard oxygen supplementation among 776 immunocompromised patients.

FLORALI was include in a 2020 meta-analysis.

Guidelines

Canadian Critical Care Trials Group, Noninvasive Ventilation Guidelines, 2011

We make no recommendation about the use of NIPPV in patients who have acute lung injury, because of a lack of RCTs.

We make no recommendation about the use of NIPPV in patients who have severe community-acquired pneumonia and no prior history of COPD

We recommend the use of NIPPV in patients who have a severe exacerbation of COPD (Grade 1A).

We recommend the use of NIPPV or CPAP in patients who have cardiogenic pulmonary edema and respiratory failure in the absence of shock or ACS requiring urgent revascularization (Grade 1A).

Design

• Multicenter, open label, randomized, controlled trial
• N=310
  • High-flow oxygen therapy (n=106)
  • Standard oxygen therapy (n=94)
  • Noninvasive ventilation (n=110)
• Setting: 23 ICU centers in France and Belgium
• Enrollment: February 2011 to April 2013
• Complete follow-up at 90 days
• Analysis: Intention-to-treat
• Primary outcome: Proportion of patients intubated at day 28

Population

Inclusion Criteria

• RR ≥ 25, AND
• PaO2:FiO2 ≤ 300 mmHg, AND (while patient breathing O2 ≥ 10L/min)
• PaCO2 ≤ 45 mmHg, AND
• No clinical history of chronic respiratory failure

Exclusion Criteria

• Cardiogenic pulmonary edema
• Severe neutropenia
• Hemodynamic instability
• Contraindication to NIPPV (e.g. low GCS)
• Urgent need for intubation

Baseline Characteristics

“Similar in all three groups”

• Age: 60
• 64% caused by community-acquired pneumonia
• 77% PaO2:FiO2 ≤ 200 mmHg
• Mean FiO2: 0.65
• SAPS II: 25

Interventions

• Randomized to one of three groups:
  • High-flow oxygen therapy: Continuous oxygen through heated humidifier and large binasal prongs, with 50L/min flow at 1.0 FiO2, adjusted to goal: SpO2 ≥ 92%
  • Standard oxygen therapy: Continuous nonrebreather face mask, ≥ 10 L/min, Goal: SpO2 ≥ 92%
  • Noninvasive ventilation: Face mask connected to ventilator with pressure support adjusted to 7-10 ml/kg (PBW) and PEEP 2-10 cm H2O, with adjustment of PEEP/FiO2 to maintain SpO2 ≥ 92%
• Minimum of 2 calendar days on therapy before switch to standard oxygen therapy allowed
• Between noninvasive ventilation sessions, patients received high-flow oxygen

Outcomes

Comparisons are high-flow vs. standard oxygen vs. noninvasive ventilation.

Primary Outcomes

Proportion of patients intubated at day 28

38% vs. 47% vs. 50% (P=0.18)

Standard oxygen vs. High-flow: OR 1.45 (0.83–2.55)

Noninvasive ventilation vs. High-flow: OR 1.65 (0.96–2.84)

Secondary Outcomes

All-cause mortality in the ICU

11% vs. 19% vs. 25% (P=0.047)

Standard oxygen vs. High-flow: OR 1.85 (0.84–4.09)

Noninvasive ventilation vs. High-flow: OR 2.55 (1.21–5.35)

All-cause mortality at 90 days

12% vs. 23% vs. 28% (P=0.02)

Standard oxygen vs. High-flow: OR 2.01 (1.01–3.99)

Noninvasive ventilation vs. High-flow: OR 2.50 (1.31–4.78)

Number of ventilator-free days at day 28

24 vs. 22 vs. 19 days (P=0.02)

Subgroup Analysis

Patients with PaO2

FiO2 ≤ 200 mmHg:

Intubation rate: 35% vs. 53% vs. 58% (P=0.01)

Number of ventilator-free days: 24 vs. 22 vs. 18 days (P<0.001)

Cause of death

Refractory shock: 46% vs 55% vs 58% (P=0.04)

Dyspnea

Marked improvement – 22% vs. 7% vs. 14%

Slight improvement– 54% vs. 34% vs. 44%

Adverse Events

No significant difference between groups

Number of cardiac arrests occurring before intubation: 2 vs. 1 vs. 0

Criticisms

• Since the majority (>80%) of these patients had pneumonia, it may be inaccurate to generalize these results to sepsis and other causes of acute hypoxemic respiratory failure without hypercapnia.
• This study was underpowered to detect or reject more subtle differences (<20%) in intubation rates.
• Patients in the noninvasive ventilation group received high-flow oxygen between treatments, which may not accurately represent the realistic comparison. Further, the two therapies may have an antagonistic relationship when used in combination.

Funding

Funded by the Programme Hospitalier de Recherche Clinique Interrégional 2010 of the French Ministry of Health. Equipment donated by Fisher and Paykel Healthcare who played no other role in the study.