Web site designed and maintained by Chris Cairns  © SICS EBM Group 2004                                  

Non-invasive ventilation following extubation in high-risk patients reduces respiratory failure but not re-intubation rate.

Citations: Early Non-invasive Ventilation Averts Extubation Failure in Patients at Risk. A Randomised Controlled Trial. Ferrer M, et al. Am J Respiratory and Critical Care Medicine. Jan 2006. Vol 173.pp 164-170

Lead author's name and fax: Miquel Ferrer. E-mail: miferrer@clinic.ub.es

Three-part Clinical Question: After extubation in patients at risk, does the early, routine use of non-invasive ventilation reduce the incidence of respiratory failure and subsequent reintubation compared to standard therapy alone?

Search Terms: Non-invasive ventilation; Extubation; Randomised controlled trial

The Study: Non-blinded randomised controlled trial with intention-to-treat.

The Study Patients: Included: patients intubated for ≥ 48 h who were extubated following a spontaneous breathing trial. At risk if ≥ 1 of  i) Age > 65, ii) Cardiac failure the cause of intubation, iii) APACHE II >12 on day of extubation.    Respiratory failure following extubation: 1) respiratory acidosis (pH ≤ 7.35 with PaCO2 >=45 mm Hg); (2) SaO2 < 90% or PaO2 < 60 mm Hg with FiO2 ≥ 0.5; (3) resp rate > 35 min–1; (4) decreased consciousness, agitation, or diaphoresis; and (5) clinical signs suggestive of respiratory muscle fatigue and/or increased work of breathing, such as the use of respiratory accessory muscles, paradoxic motion of the abdomen, or retraction of the intercostal spaces. 'Rescue' non-invasive ventilation was used in both groups in the case of respiratory failure after extubation without the need for immediate re-intubation. In the NIV group this was continued for 24 hours, but in the standard therapy group this was discontinued within 4 hours, and if still required the patient was re-intubated. Nine patients from the control group and four from the NIV group escaped re-intubation due to rescue NIV. 51% of the patients had COPD.
 

Control group (N = 83; 83 analysed): Post extubation, patients were given oxygen via a venturi mask. The inspired oxygen fraction titrated to achieve a saturation >92%.
 

Experimental group (N = 79; 79 analysed): Immediately after extubation non-invasive ventilation was continuously delivered using bi-level positive airway pressure mode, using a tightly fitting facemask as the first choice of administration. Expiratory airway pressure was set at 4-6cmH2O and a FiO2 was titrated to achieve a SpO2>92%. NIV was administered for as much time as possible for the first 24hours after extubation.

The Evidence:

 EBM Questions:

1. Do the methods allow accurate testing of the hypothesis? Yes.   Although, the use of 'rescue' NIV in both groups may underestimate the true extent of the benefit of NIV routinely. The exact NIV treatment protocol is vague, with the description of its use 'for as much time as possible over a 24 hour period'.   No time periods were documented.

2. Do the statistical tests correctly test the results to allow differentiation of statistically significant results? Yes

3. Are the conclusions valid in the light of the results? Yes.

4. Did results get omitted and why? No. All patients were followed up whether they required immediate or later reintubation.

5. Did they suggest areas of further research? Yes. Due to the high proportion of patients included in this trial with chronic respiratory disorders and hypercapnoea, they suggest extrapolating these findings and applying them solely to those patients with hypercapnoea with chronic respiratory disorders.

6. Did they make any recommendations based on the results and were they appropriate? No

7. Was this study relevant to my clinical practice? Yes

8. What level of evidence does this represent? 1⁺

9. What grade of recommendation can I make from this result alone? B

10. What grade of recommendation can I make from this study when considering all the available evidence? B. Interestingly: There was a high incidence of respiratory failure in the control group compared with previous studies (33% vs. 25%). Time from extubation to re-intubation did not influence mortality which differed from most previous studies looking at this.

11. Should I change my practice because of these results? No.   We give more weight to trials were a clinical meaningful endpoint was altered.   The primary endpoint in this study was “respiratory failure”. While we all recognise the definition used, we would be more impressed by impact on re-intubation rate, length of ICU stay and hospital mortality.   While the study did show a reduction in ICU mortality this was out of keeping with the other results and is larger than we would expect.   The mortality benefit was not sustained to hospital discharge.

12. Should I audit my current practice because of these results? Yes: What proportion of your patients would be classified as high-risk?  What is your re-intubation rate?

Appraised by: Pam Doherty SpR department of Anaesthesia and Intensive Care, Western Infirmary, Glasgow; 13 June 2006
Email: pamdoherty@doctors.org.uk

Kill or Update By: June 2009

Edited by CC & MD

Citation: EBM Critical Appraisals. Scottish Intensive Care Society EBM Group. 2006. Doherty P. Early Non-invasive Ventilation Averts Extubation Failure in Patients at Risk. A Randomised Controlled Trial. Ferrer M, et al. Am J Respiratory and Critical Care Medicine. Jan 2006. Vol 173.pp 164-170

 Printer friendly view

 ©SICS EBM 2006