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

Broncho-alveolar Lavage v. Endotracheal Aspiration in the Diagnosis of Ventilator Associated Pneumonia

Citation: A Randomised Trial of Diagnostic Techniques for Ventilator-Associated Pneumonia.
The Canadian Critical Care Trials Group; NEJM 2006, 355: 2619 – 2630

Lead author: Dr D Heyland; dkh2@post.queensu.ca

Three-part Clinical Question:

Patient: ICU patients with suspected Ventilator Associated Pneumonia (VAP)

Interventions: Quantitative culture of bronchoalveolar lavage fluid v. Non-quantitative culture of endotracheal aspirate

Outcomes: Primary: 28-day mortality; Secondary: length of stay in ICU and hospital, duration of mechanical ventilation, response to clinical and microbiological treatment and use of antibiotics

Search Terms: Ventilator Associated Pneumonia; Canadian trials

The Study: Multi-centre, non-blinded, concealed, randomised trial with intention-to-treat.

The trial had a 2x2 factorial design. The factors were invasive v. non-invasive diagnostic techniques and monotherapy v. combination empirical antibiotics. The diagnostic techniques were considered in this paper.

The Study Patients: 740 patients from ICUs in Canada and the United States. All patients are immunocompetent adults with suspected VAP after 4 days in the ICU. 

Empirical antibiotic administration was standardized in all study patients. Once the culture result was available: a negative culture resulted in cessation of the study antibiotics (except in high risk VAP patients at the discretion of the attending physicians); a positive culture resulted in a change to a single narrow spectrum antibiotic according to local practices.

Control group: (N = 374; 374 analysed): Patients who had endotracheal aspiration with non-quantitative culture of endotracheal aspirate.

Experimental group: (N = 365; 365 analysed): Patients who had bronchoalveolar lavage with quantitative culture of bronchoalveolar lavage fluid

One patient withdrew consent 2 days after randomization and data for that patient were not analyzed further.

The Evidence:

¶ Targeted therapy was defined as the discontinuation or modification of study antibiotics on the basis of the organisms

cultured or the resumption of antibiotics to treat a pre-enrollment condition if the culture was negative.

* time from randomisation

 Comments:
The findings suggest that either diagnostic approach is acceptable in the presence of adequate initial empirical antibiotic cover and targeted therapy based on culture results in VAP.
Of the initial pool of 2531 patients, 1013 (40%) were excluded. They were immunocompromised, colonized or infected with Pseudomonas or methicillin-resistant Staphylococcus aureus; allergic to, or had recently received, one of the study drugs; likely to die within 72 hrs of enrollment and likely to remain on ICU for more than three weeks. It can be argued that patients more likely to benefit from an invasive diagnostic strategy were excluded from the trial.

The trial protocol also allowed physicians to treat patients with prior antibiotic exposure who had <10⁴ CFUs on bronchoalveolar lavage fluid. It is not stated how many patients in the experimental group were treated thus. It could be argued that the trial may have compared non-quantitative diagnostic techniques.  The primary outcome of 28 day mortality assumes a direct relationship between diagnostic techniques and survival, disregarding the many other variables that contribute to mortality in the intensive care patient.

 EBM questions:

1) Do the methods allow the adequate testing of the hypothesis?

Generally yes, but at least 40% of the screened patients who were excluded had risk factors for colonization or infection with potentially antimicrobial-resistant bacteria. Use of BAL may be advantageous in these high-risk patients; an accurate microbiological diagnosis and more appropriate use of antibiotics may improve outcome.

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

3) Are conclusions valid in light of results?

Perhaps, it depends on whether this trial actually compared the intended diagnostic techniques (see above comment)

4) Did results get omitted, and why?   Yes, though only from one patient.

5) Did they suggest areas of further research?   No

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

7) Is this study relevant to my clinical practice?

Yes, in so far as it reports the absence of a significant advantage of one diagnostic method over the other, when used in conjunction with broad spectrum empirical antibiotics. However when assessing applicability to other patients we must consider the large number and type of excluded patients. The findings may not be applicable to patients with risk factors for colonisation or infection with potentially antimicrobial-resistant bacteria.

8) What level of evidence does this study represent?  1^-

9) What grade of recommendation can I make on this result alone?

It is difficult to apply our grading system to this study.

10) What grade of recommendation can I make when this study is considered along with
other available evidence?

Again, it is difficult to make a recommendation given the different designs of such studies.

 
11) Should I change my practice because of these results?

No, since the study did not prove any significant advantage of one diagnostic technique over the other. Any potential advantage of BAL over endotracheal aspiration for accurate microbiological diagnosis and subsequent cessation of broad-spectrum antibiotic therapy in patients who likely to have resistant bacteria remains unproven as this study excluded these patients. If subsequent studies in this population produce similar results then there would be a significant financial incentive to discontinue quantitative measurements.

12) Should I audit my current practice because of these results?

No, since audit is a tool used to compare ones current practice with evidence based gold standard. This paper fails to provide evidence of a gold standard diagnostic technique for VAP.

Appraised by: Dr Srikanth Chukkambotla, SpR, Dr Imelda Galvin, SpR, Department of Anaesthetics & Intensive Care, Royal Oldham Hospital, UK; 20 January 2007
Email: srikanth73@rediffmail.com

References:

1.      A Randomised Trial of Diagnostic Techniques for Ventilator- Associated Pneumonia.
The Canadian Critical Care Trials Group; NEJM 2006; 355: (25) 2619 – 2630

2.      Kollef, M. H.  Diagnosis of Ventilator-Associated Pneumonia. NEJM 2006; 355: 2691-2693

Reviewed & edited by Chris Cairns & David Swann.

Citation: EBM Critical Appraisals. Scottish Intensive Care Society EBM Group. 2007 & JICS 2007 Vol8(1). Chukkambotla S, Galvin I. A Randomised Trial of Diagnostic Techniques for Ventilator-Associated Pneumonia.
The Canadian Critical Care Trials Group; NEJM 2006, 355: 2619 – 2630

Printer friendly view

©SICS EBM 2007