Humidified Non-Invasive Ventilation versus High-Flow Therapy to Prevent Reintubation in Patients with Obesity

Hernández et al. AJRCCM 2025. DOI: 10.1164/rccm.202403-0523OC

Clinical Question

• In obese adult patients at intermediate risk for hypoxaemic extubation failure, does non-invasive ventilation (NIV) therapy with active humidification compared to use of high-flow nasal oxygen (HFNO or HFNC) reduce all-cause reintubation within 7 days after extubation?

Background

• One in eight people worldwide are considered to be obese
• Obesity is associated with impairment of normal respiratory physiology, including increased intra-abdominal and pleural pressures, atelectasis, obstructive sleep apnoea (OSA), and obesity hypoventilation syndrome (OHS)
• Post-extubation respiratory failure results in prolonged ICU and hospital stays, greater costs to the healthcare system, and increased patient mortality
• The optimal post-extubation respiratory support is currently unknown, however NIV is thought to be beneficial in obesity due to modifying the above impairments in respiratory function
• There are currently conflicting results in studies on weaning patients from mechanical ventilation, particularly in the setting of obesity
• Prior studies assessing respiratory support post-extubation or for hypoxaemic respiratory failure:
  • Thille et al. 2024 – post-hoc analysis of three large-scale randomised controlled trials assessing the effect of prophylactic NIV after extubation in obese patients, all studies suggested that NIV is beneficial but not always statistically significant
  • Thille et al. 2019 (HIGH-WEAN) – n=641, HFNC alternating with NIV reduced risk of reintubation at day 7 compared to HFNC alone
  • De Jong et al. 2023 (EXTUB-OBESE) – n=981, NIV reduced relative risk of extubation failure compared to oxygen therapy, though most of the difference was due to patients in the oxygen therapy group switching to NIV
  • Stéphan et al. 2015 (BIPOP) – n=830, high-flow nasal oxygen therapy was not inferior to BiPAP in patients with respiratory failure post cardiothoracic surgery

Design

• Multicentre randomised controlled superiority trial
• Recruitment:
  • Screened for extubation readiness daily
    • Recovery from precipitating illness
    • Respiratory criteria – PF ratio >150, PEEP <8, pH >7.35
    • Clinical criteria – no/low vasoactive drugs, HR <140, Hb >80, temp <38, no need for sedatives, presence of respiratory stimulus, appropriate spontaneous cough
  • If meeting criteria for extubation readiness, then patients underwent a spontaneous breathing trial (SBT)
    • Trial of spontaneous ventilation mode with pressure support of 7 cmH₂O for 30 minutes
    • SBT failure criteria included:
      • Agitation, anxiety, depressed mental state, diaphoresis, cyanosis, increasing respiratory effort, facial signs of distress
      • Respiratory – dyspnoea, PaO2 <60 or SpO2 <90% on FiO2 >0.5, PaCO2 >50 or increased >8 from baseline, pH <7.32 or decreased >0.07 from baseline, RR >35 or >50% from baseline
      • Cardiovascular – HR >140 or increased >20% from baseline, SBP >180 or >20% from baseline, SBP <90, arrhythmias
  • If SBT was tolerated, then patients were reconnected to the ventilator with the previous settings until planned for extubation
  • Patients who passed the SBT were randomised (simple randomisation) to receive NIV or HFNC by means of a phone call to a call centre immediately before planned extubation after reconnection for rest
  • Written informed consent was obtained from the patient’s relative on the day of the first SBT
• Defined clinical protocols were used for titration of NIV and HFNC
• Data was collected from bedside nursing charts, medical notes, pathology reports, electronic monitors, and interviews with relatives
  • Continued until final discharge from hospital
• Analysed on intention-to-treat basis
• Sample size of 144 patients with 72 patients in each treatment arm
  • Based on estimated absolute reduction in reintubation rate of 14% (from baseline of 17%) in order to achieve 80% power to detect difference with ɑ level of 5% and maximum tolerated patient loss rate of 10%
    • Estimated reintubation rate in NIV group was therefore 3%
• Open-label and clinicians were unable to blinded due to the nature of the intervention

Setting

• Conducted in two ICUs in Spain
• Data collected from June 2020 to June 2021

Population

• Inclusion:
  • • Adult patients receiving mechanical ventilation for >24 hours and deemed ready for scheduled extubation according to tolerance of a spontaneous breathing trial, who have a BMI ≥30 kg/m² and fulfil ≤2 of following criteria for high-risk for extubation failure:
      • Age >65
      • Heart failure as primary indication for mechanical ventilation
      • Moderate to severe COPD
      • APACHE II >12 points on extubation day
      • Airway patency problems (including high risk of developing laryngeal oedema)
      • Inability to deal with respiratory secretions (inadequate cough reflex or suctioning >2 times within 8 hours before extubation)
      • Difficult or prolonged weaning
      • 2 or more comorbidities (Charlson Comorbidity Index)
      • Prolonged (>7 days) mechanical ventilation
    • Only the first extubation episode was assessed
• Exclusion:
  • • Age <18 years
    • Pregnancy
    • Do-not-resuscitate orders
    • Tracheostomy prior to extubation
    • Accidental extubation or self-extubation
    • Hypercapnoea (P_aCO₂) >45mmHg at end of spontaneous breathing trial
• 2,356 patients assessed for eligibility –> 634 eligible –> 144 randomised
  • 72 patients randomised to receive NIV and 72 patients randomised to receive HFNC
• There were some imbalances in patient characteristics, some of which favoured the NIV group and some of which favoured the HFNC group in terms of reintubation risk
• Comparing baseline characteristics of NIV vs. HFNC group:
  • Median age: 62 vs. 60 years
  • Women: 33% vs. 32%
  • Median APACHE II score: 16.9 vs. 15.9
  • Median length (days) of MV before extubation: 6 vs. 3
  • BMI: 33 vs. 33
  • Comorbidities:
    • Heart disease 19% vs. 17 %
    • COPD 10% vs. 3%
  • High-risk factors for reintubation:
    • Age >65: 32% vs. 28%
    • Heart failure as primary indication for MV: 4% vs. 1%
    • COPD: 7% vs. 3%
    • APACHE II score >12 on extubation day: 8% vs. 3%
    • Airway patency problems: 6% vs. 25%
    • Inability to deal with respiratory secretions: 7% vs. 17%
    • Difficult or prolonged weaning: 19% vs. 15%
    • Two or more comorbidities: 45% vs. 60%
    • Prolonged MV (>7 days): 32% vs. 25%
    • Median number of high-risk factors: 3 vs. 3
  • Diagnosis at admission:
    • Respiratory primary failure: 71% vs. 36%
    • COVID-19: 44% vs. 24%
    • Haemodynamic failure: 39% vs. 35%
    • Neurologic failure: 40% vs. 53%
    • Surgical: 22% vs. 33%

Intervention

• NIV with active humidification for 48 hours post extubation, then changed to conventional oxygen therapy
  • Started immediately after extubation
  • Continuously delivered by bilevel positive airway pressure through size-fitted face mask
  • Pressure support and positive end-expiratory pressure were adjusted to target RR 26 and adequate gas exchange based on SpO₂ and arterial blood gases
  • HFNC as alternative therapy was not allowed
  • Sedatives to increase tolerance were not allowed
    • Median IPAP 4 cm H₂O, Median EPAP 8 cm H₂O (Table E3)

Control

• HFNC for 48 hours post extubation, then changed to conventional oxygen therapy
  • Started immediately after extubation
  • Flow initially set at 10L/min then rapidly titrated upward in 5 L/min increments until patients experienced discomfort or up to 60L/min
  • NIV as rescue therapy not allowed
    • Median flow 60L/min

Management common to both groups

• Both groups were treated by the same medical, nursing, and respiratory therapy staff
• Both groups received similar medical management
  • Details were not specified in the article or data supplement

Outcome

• Primary outcome:
• All-cause reintubation within 7 days after extubation:
  • There was no significant difference in reintubation rates between the NIV and HFNC groups
  • 17 patients (23.6%) in NIV group vs. 24 patients (33.3%) in HFNC group required reintubation
    • Absolute risk difference 9.7% (95% CI -4.9 to 24.4)
    • When adjusted for risk factors related to increased risk for extubation failure in patients with obesity, no significant benefit with NIV was observed (OR 0.71, 95% CI 0.32-1.58)
• Secondary outcomes:
  • No significant difference in
    • Post-extubation respiratory failure
    • Ventilator-associated pneumonia
    • Hospital LOS
    • ICU and hospital mortality
    • Time to reintubation
  • Significantly greater in NIV group:
    • ICU LOS
      • Median 11 days in NIV group vs. 6.5 days in HFNC group (p=0.02)
    • Intolerance to therapy
      • 15% in NIV group vs. 4% in HFNC group (p=0.049)
• Post Hoc Bayesian Analysis:
  • suggested high probability of benefit in NIV reducing reintubation compared to HFNC:
    • Data-driven prior – RR 0.65 (95% CrI 0.44-0.96), posterior probability of RR <1 = 0.99
    • Minimally informative prior – RR 0.70 (95% CrI 0.37-1.30), posterior probability of RR <1 = 0.87

Authors’ Conclusions

• Among obese adult patients with intermediate risk of extubation failure, treatment with NIV did not significantly reduce the risk of reintubation within 7 days compared to HFNC

Strengths

• Allocation concealment via phone call to a call centre
• Assessed the reintubation rate at 7 days compared to many previous studies that assessed reintubation rate at 3 days
  • Important given that morbidity and mortality has historically been highest when reintubation is required at 72-96 hours post extubation, and also that approximately ~25% had a time to reintubation > 72 hours in NIV group based on median time to reintubation was 52 (IQR 27 – 74) hours
• Detailed clinical protocol for starting, titrating, and modifying NIV and HFNC post extubation
  • Increases ability of this study to be replicated in other settings
• High adherence to therapy in first 48 hours
• Interventions are easy to implement in all centres where the equipment is available
• Complete follow up with no crossover between NIV and HFNC groups
• No evident conflicts of interest including financial support

Weaknesses

• The study was conducted in two ICUs in Spain which may not be representative of all places of practice
• Median PEEP of 8cm H20 in NIV group – given physiological rationale of PEEP in this context, could higher levels of PEEP improve outcomes, or potentially lead to worse tolerance?
• Included a large number of patients with COVID-19 at admission (49 out of 144)
  • Results might become less applicable to the general ICU population as rates of respiratory failure secondary to COVID-19 decline
• Open-label and unable to be blinded due to nature of intervention
  • Increases risk of bias as nursing, medical, and respiratory therapy staff are aware of which treatment arm patients are assigned to which could impact clinical decision making for reintubation
  • Will be an innate problem with all studies investigating similar interventions
• Imbalance in certain baseline characteristics between groups:
  • Contributed to by small sample size and simple randomisation of groups
  • NIV group had greater proportion of patients with longer mechanical ventilation, COPD, COVID-19, and primary respiratory failure
  • HFNC group had greater proportion of patients with airway patency problems, neurologic failure, trauma, and surgical pathologies
• Absolute risk reduction of 14% used in power calculation seems overly optimistic
  • Based on previous studies (including those linked above), the absolute risk reduction with NIV is likely to be lower than 14%

The Bottom Line

• Pragmatic study that investigates an important clinical question in the ICU
• Whilst the results from this study were not statistically significant, given the signal of benefit seen in other studies (and the post-hoc Bayesian analysis) combined with the fact that NIV has been shown to be a fairly low-risk intervention and reasonably well tolerated I will continue to use NIV prophylactically in a select group of obese patients

External Links

• Article Humidified Noninvasive Ventilation versus High-Flow Therapy to Prevent Reintubation in Patients with Obesity: A Randomized Clinical Trial
• Editorial Prophylactic Noninvasive Ventilation after Extubation of Obese Patients

Metadata

Summary author: Sarah Scholz
Summary date: 19th March 2025
Peer-review editor: George Walker

Picture by: Pixabay/Pexels