Critical Care Alert, Critical Care, Airway

CCA: Effect of NIV after Extubation in Critically Ill Patients with Obesity

ARTICLE: De Jong A, Bignon A, Stephan F, et al. Effect of non-invasive ventilation after extubation in critically ill patients with obesity in France: a multicentre, unblinded, pragmatic randomised clinical trial. Lancet Respir Med. 2023;11(6):530-539.


The EXTUB-OBESE study compared non-invasive ventilation (NIV) vs oxygen therapy (high-flow nasal oxygen [HFNO] or standard oxygen) after extubation of critically ill patients with obesity to see which therapy reduced treatment failure (reintubation or switch to other treatment)


The best post-extubation strategy in critically ill patients with obesity is currently unknown. Avoiding reintubation is key as it is associated with increased mortality. Reintubation rates after extubation may exceed 20% in patients at high risk of extubation failure.1 Individuals with obesity have challenging physiology as they have a lower functional residual capacity, higher oxygen demand, are susceptible to obstructive apnea, and have a higher transthoracic pressure, making them more susceptible to atelectrauma. This puts these patients at a higher risk both during intubation and extubation. Atelectasis persists after extubation more in obese patients than non-obese ones.2 NIV combats this additional load by adding positive transpulmonary pressure, which may lead to more successful extubations. Another popular extubation strategy is to extubate to HFNO which provides a lower positive pressure than that capable by NIV but still provides augmentation. Finally, many patients are extubated to low-flow nasal cannula only.

Previous studies on HFNO vs. NIV in obese patients have been mixed. A post-hoc analysis of thoracic surgery patients did not show a superiority of NIV.3 An additional post-hoc analysis showed a significantly decreased extubation failure rates when using NIV and HFNO vs HFNO only after stratification by BMI.4 Another study comparing high-flow nasal cannula (HFNC) to standard oxygen in obese patients did not show a statistical significance. None of these studies compared in the same RCT all of the oxygen therapy devices available in specifically obese patients that received mechanical ventilation for various indications. This trial was designed to compare the results of extubating to high-pressure NIV with oxygen therapy vs HFNO and oxygen therapy in obese patients. The trial further looked at effects of alternating NIV with HFNO or standard oxygen in the NIV group.


This study was a multicenter, parallel-group, unblinded, pragmatic randomized trial comparing prophylactic NIV applied immediately after extubation alternating with HFNO or standard oxygen vs. use of oxygen therapy alone (HFNO or standard oxygen). The trial took place from October 2019 to July 2021 and was conducted in 39 ICUs in France.

Participants underwent two randomizations. The first central randomization decided whether patients would be in the NIV or oxygen therapy group. The second randomization determined the method of oxygen administration in each group (HFNO or standard oxygen). Randomization was done using a blinded assignment sequence, stratified by length of mechanical ventilation, type of admission, and by the center. Intervention blinding was not possible. The assignments were blinded from the statisticians on data analysis.

In the NIV group, patients were put on NIV within 30 minutes of extubation. PEEP values were set to 10 cm H2O while pressure support was set to obtain a rate between 20-30 breaths/min and an expired tidal volume between 6-8 mL/kg of predicted body weight.  Patients underwent NIV sessions of 30-60 minutes spread through the day and night for at least 4 hours in the day with no upper limit in the first 24 hours. In both groups, HFNO was administered with a flow rate of 50 L/min at a FiO2 set to target oxygen saturation >94% in both groups. After 24 hours, NIV and HFNO were continued if the patient still needed additional support. The follow-up was stopped after 3 months.

Analysis was an intention-to-treat comparison among patients in the two trial groups with use of an uncorrected chi-square test. Further subgroup analysis was completed among medical vs. surgical patients, length of ventilation, and a subgroup with SarsCOV2 infection. Additional post-hoc primary outcome analysis was completed with a per-protocol analysis and crossover analysis for patients that switched from oxygen therapy group to the NIV group.


Patients older than 18 years of age, admitted to the ICU, and covered by public health insurance.

Patients were included if they met criteria for extubation in the ICU after being on mechanical ventilation for >6 hours and had obesity as defined by body-mass index (BMI) of 30 kg/m2 or greater on the day of extubation.


Patients were excluded if they had hypercapnia (PaCO2 ≥ 50 mm Hg) before extubation (which was a mandatory indication for NIV after extubation), isolated cardiogenic pulmonary edema, tracheotomy, home ventilation (BiPAP or CPAP), patients with decision of “do not reintubate,” anatomical factors restricting use of NIV of HFNO,  or previous extubation during ICU stay with previous inclusion in the study.


The primary outcome of the study was the failure rate within three days after extubation, defined as any of the following: reintubation for mechanical ventilation, switch to the other study treatment, or premature study-treatment discontinuation (at the request of the patient or for medical reasons such as gastric distension). Switch to the other study treatment was considered in the use of other treatments as a rescue therapy to avoid reintubation.


The secondary outcome of the study was the incidence of acute respiratory failure within 7 days after extubation.

Additional exploratory clinical outcomes were evaluated: oxygenation evaluated by the ratio of pressure of  PaO2 to FiO2 until day 7, organ failure assessed with SOFA scores until day 7, reintubation rates within 7, 14, and 28 days after extubation, length of stay in ICU and in hospital, ICU mortality rate, day 28 and day 90 mortality rates.


The final cohort contained 981 individuals who met the inclusion criteria. 490 were assigned to the NIV group and 491 were assigned to the oxygen therapy groups.  The characteristics were balanced between the two treatment groups.

Primary outcome

In the intention-to-treat analysis, 66 (13%) of the 490 patients in the NIV group had treatment failure within 3 days vs 130 (26.5%) of the 491 patients in the oxygen therapy group (relative risk 0.43; 95% CI 0.31-0.60, p<0.0001). Neither the method of oxygen delivery while on oxygen therapy in the NIV, nor the other subgroup analysis modified the effect of the NIV group on treatment failure rate. The results of the per-protocol analysis were consistent with the intention-to-treat analysis.

Secondary outcomes
Fifty-four (11%) of the 490 patients in the NIV group vs. 70 (14%) of the 491 patients in the oxygen therapy group experienced acute respiratory failure within 7 days after extubation (relative risk 0.75; 95% CI 0.51 to 1.09, p=0.13). These results were not statistically significant. The delay between extubation and respiratory failure nor the cause of respiratory failure did not differ between the two groups.

Further exploratory outcomes did not show a significant difference including need for reintubation within 7, 14, and 28 days after extubation, ICU or hospital length of stay, ICU mortality at the 28 or 90-day mark, or the PaO2/FiO2 ratio and SOFA scores at the 7-day mark.


  • Although the study was performed across 39 centers, approximately 50% of enrolled patients came from 6 centers. Thus the results may not be generalizable across all hospitals.
  • There was a lack of systematic spontaneous-breathing-trials performed prior to extubation.
  • The implementation of the interventions in the study were unblinded.
  • Specific types of acute respiratory failure or specific procedures prior to intubation were not recorded.
  • NIV sessions were designed to be shorter and spaced. These were intended to be preventative rather than curative. Longer sessions of NIV may have further improved outcomes.


The EXTUB-OBESE study showed a decrease in extubation failure rates when extubating to NIV for obese patients vs HFNO/standard oxygen therapy. An extubation plan with NIV should be strongly considered for any obese patient and may be cycled with oxygen therapy for comfort. Although we may extubate less frequently in ED, when doing so, it is important to consider that obese patients have additional risk factors that make NIV an attractive option that may decrease the need for reintubation.


  1. Thille AW, Richard JC, Brochard L. The decision to extubate in the intensive care unit. Am J Respir Crit Care Med. 2013;187(12):1294-1302.
  2. Dixon AE, Peters U. The effect of obesity on lung function. Expert Rev Respir Med. 2018;12(9):755-767.
  3. Stéphan F, Bérard L, Rézaiguia-Delclaux S, Amaru P; BiPOP Study Group. High-Flow Nasal Cannula Therapy Versus Intermittent Noninvasive Ventilation in Obese Subjects After Cardiothoracic Surgery. Respir Care. 2017;62(9):1193-1202.
  4. Thille AW, Coudroy R, Nay MA, et al. Beneficial Effects of Noninvasive Ventilation after Extubation in Obese or Overweight Patients: A Post Hoc Analysis of a Randomized Clinical Trial. Am J Respir Crit Care Med. 2022;205(4):440-449.

Related Articles

Critical Care Alert: Hydrocortisone plus Fludrocortisone for Adults with Septic Shock (APROCCHSS)

Can the use of hydrocortisone plus fludrocortisone and/or drotrecogin alfa improve clinical outcomes in patients with septic shock?

Understanding Alcoholic Ketoacidosis

Understanding Alcoholic Ketoacidosis A 49-year-old male with a history of alcohol abuse presents to the ED with complaints of generalized abdominal pain and vomiting for the last 36 hours. The patien