Hernández

HFNC

Effect of Postextubation High-Flow Nasal Cannula vs Conventional Oxygen Therapy on Reintubation in Low-Risk Patients. A Randomized Clinical Trial

Hernández. JAMA. 2016; Published online March 15, doi:10.1001/jama.2016.2711

Clinical Question

  • In patients that are low risk for re-intubation does high flow nasal oxygen compared with conventional oxygen therapy reduce re-intubation within 72 hours?

Design

  • Randomised controlled trial
    • Use of random number generator through telephone call centre
    • Permuted blocks of 10
    • Randomisation stratified by hospital
  • Follow up until discharge from hospital
  • Intention to treat analysis
  • Blinding
    • Clinicians and patients non-blinded
    • Statistical analysis performed blinded
    • Investigators did not participate in clinical decisions
  • Statistics
    • Sample size calculation: 260 patients would give a 80% power to detect an absolute risk reduction of 8% from a baseline of 13%, with a false positive rate of 5%
    • Primary outcome compared with Cochran-Mantel-Haenszel χ2 test

Setting

  • 7 Intensive Care Units in Spain
  • Trial conducted from September 2012 – October 2014

Population

  • Inclusion criteria:
    • Adult patients who had been mechanically ventilated for at least 12 hours and ready for scheduled extubation following a spontaneous breathing trial
    • Low risk for re-intubation:
      • Age <65; APACHE II score on day of extubation <12; BMI <30; adequate secretions management; simple weaning; <2 co-morbidities; mechanical ventilation <7 days; absence of: heart failure, moderate-to-severe chronic obstructive pulmonary disease, airway patency problems
  • Exclusion criteria: Do-not-resuscitate order; tracheostomy; accidental extubation; hypercapnic during spontaneous breathing trial
  • 527 patients randomised out of 10347 patients who were screened.
  • 8608 patients excluded as ‘not weanable’

Comparing intervention vs. control groups

  • Age (mean): 51.1 vs. 51.8
  • APACHE II:
    • at ICU admission (median): 14 vs. 13
    • at extubation (median): 7 vs. 7
  • Length of mechanical ventilation before extubation, median days (IQR): 1 (1-3) vs. 2 (1-4)
  • Co-morbidities:
    • Neurological disease: 7.6% vs. 12.9%
    • Heart disease: 7.6% vs. 8.7%
    • Mild COPD: 3% vs. 1.9%
  • Diagnosis at admission:
    • Medical
      • Respiratory primary failure: 16.3% vs. 16.7%
      • Neurologic primary failure: 26.1% vs. 32.7%
    • Traumatic brain injury: 11.7% vs. 6.5%
    • Surgical: 49.6% vs. 45.6%
  • PaO2:FiO2 mmHg from spontaneous breathing trial prior to extubation (mean) 227 vs. 237

Intervention

  • High flow nasal oxygen for 24 hours post extubation
    • Flow initially set at 10l/min and titrated upwards in 5l/min steps until patient felt discomfort
    • Temperature initially set at 37C
    • FiO2 adjusted to target SpO2 >92%

Control

  • Conventional oxygen therapy for 24 hours post extubation
    • Applied through nasal cannula or non-rebreather facemask
    • FiO2 adjusted to maintain SpO2 >92%

Outcome

  • Primary outcome: Re-intubation within 72 hours – significantly reduced in high flow nasal oxygen group:
    • 13 patients (4.9%) vs. 32 patients (12.2%), absolute difference, 7.2% (95% CI, 2.5% to 12.2%); p = .004, Number needed to treat 14 (95% C.I. 8-40)
    • Fragility index: If 5 extra patients had required re-intubation in the high flow nasal oxygen group then the result would have no longer been statistically significant
  • Secondary outcomes – no significant difference in any of the following:
    • Comparing intervention group vs. control group
      • ICU mortality: 1.1% vs. 1.1%
      • Hospital mortality: 3.8% vs. 5%, p=0.94
      • Respiratory infections: 2.3% vs. 4.9%, absolute difference 2.7 (95% C.I. -0.6 to 6.2), p=0.07
  • Outcomes that were not pre-specified on clinicaltrials.gov – Comparing intervention vs. control groups:
    • Post extubation respiratory failure: significantly lower in high flow nasal oxygen group
      • 22 patients (8.3%) vs. 38 patients (14.4%), absolute difference 6.1% (95% C.I. 0.7-11.6), p=0.03
    • Cause of post-extubation respiratory failure: No. (%)
      • Inability to clear secretions: 3 (13.6%) vs. 14 (36.8%)
      • Hypoxia: 7 (31.8%) vs. 6 (15.8%)
      • Unbearable dyspnoea: 9 (40.9%) vs. 14 (28.9%)
    • Reasons for re-intubation
      • Non-respiratory causes: (surgery, low GCS) – no significant difference
        • 9 patients (3.4%) vs. 9 patients (3.4%)
      • Respiratory causes for re-intubation: significantly lower in high flow nasal oxygen group
        • 4 patients (1.5%) vs. 23 patients (8.7%),  absolute difference, 7.2% (95% CI, 3.6% to 11.4%) p=0.001
      • Breakdown of respiratory causes for re-intubation:
        • Persistent post extubation respiratory failure: 2 patients (0.8%) vs. 16 patients (6%)
        • Inability to clear secretions: 0 patients vs. 5 patients (1.9%)
        • Haemodynamic impairment/cardiac arrest: 1 patients (0.4%) vs. 2 patients (0.8%)
        • Agitation: 1 patient (0.4%) vs. 0 patients
      • A post-hoc analysis also analysed if patients had laryngeal oedema that required re-intubation – there were significantly greater patients in the conventional oxygen therapy group
        • 0 patients vs. 7 patients (3.1%), p=0.001
        • The primary outcome of re-intubation within 72 hours was re-analysed after patients requiring re-intubation for laryngeal oedema were excluded. There was still a significant reduction in re-intubation rates in the high flow nasal oxygen group
          • 4.9% vs 9.8%, p = .04
    • Time to re-intubation (median hours): 19 vs. 15, p=0.66
    • No adverse effects were noted
    • Compared with re-intubated patients, successfully extubated patients had shorter duration of mechanical ventilation 1 (range, 1-2) vs 3 (range, 1-3) days; absolute difference, 2 days (95% CI, −2 to 4); p < .001

Authors’ Conclusions

  • In patients that were low risk for requiring re-intubation the use of high flow nasal oxygen reduced the risk of re-intubation

Strengths

  • Randomised controlled trial
  • Allocation concealment maintained
  • Statistical analysis performed blinded
  • Zero loss to follow up
  • Intention to treat analysis
  • Standardised criteria for when to attempt extubation, and when to perform re-intubation
  • Listed on clinicaltrials.gov

Weaknesses

  • Large number of patients excluded included over 80% who were not weanable. This limits the external validity.
  • Considering these were low risk patients there were a large number of patients that required re-intubation in the control group. This may limit the external validity to centres that have a lower re-intubation rate in low risk patients.
  • There were some baseline differences between the intervention and the control group. There was a significantly lower incidence of neurological co-morbidity in the high flow nasal oxygen group. This confounding variable may have introduced bias. The length of mechanical ventilation was shorter, although this did not reach statistical significance, in the high flow group. As the authors reported that successfully extubated patients had a shorter duration of ventilation, this may have also introduced bias.
  • In clinicaltrials.gov the primary outcome was planned to be measured at 3 months, where as in the paper it was measured at 72 hours. A number of secondary outcomes were also added after the planned analysis was published on clinicaltrials.gov

The Bottom Line

  • This multi-centre randomised controlled trial found that in patients that are low risk for requiring re-intubation, the use of high flow nasal oxygen compared with conventional oxygen therapy reduced the risk of re-intubation at 72 hours. However, due to some baseline differences, and some concerns over the external validity I would want to see further trials before this becomes routine practice.

External Links

Metadata

Summary author: David Slessor
Summary date: 16th March 2016
Peer-review editor: Adrian Wong

7 comments

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.