Diagnosing Ventilator Associated Pneumonia

Todd Fraser on 17-05-2011

Ventilator associated pneumonia is an important cause of acquired morbidity in the ICU. Unlike community acquired pneumonia, the diagnosis is not always straight forward, as many causes of lung infiltrates, fever, purulent sputum and leukocytosis exist. Tracheal aspirate cultures are well recognised to be overly sensitive, with many positive cultures representing colonisation. In this context, empiric antibiotics may be over used, resulting in unnecessary side effects, cost and potentially generation of resistant organisms. The best method for diagnosing VAP has remained elusive and an area of controversy for many years. While lung biopsy remains the gold standard, it is associated with morbidity, and efforts have centred on less invasive methods. Unfortunately, direct comparisons between the gold standard an alternative strategies are lacking. Early evidence suggested however that bronchoscopic, quantitative methods such as Bronchoalveolar Lavage (BAL) and Protected Brush Specimens (PBS) could be safely used to limit antibiotic therapy (ie, if the BAL or PBS was negative, and antibiotics were withheld, patients did no worse). This has been seen as a benefit. Bronchoscopy however carries the burden of requiring specific skills which may not be available 24 hours a day. "Blind" methods appear to be just as sensitive. So does quantitative vrs qualitative make any difference? Or deep sampling vrs tracheal aspirate make any difference? According to a 2008 Cochrane review, there is little evidence to support either approach, with mortality, length of stay, duration of ventilation and changes to initial antibiotic therapy all similar. So given the ease, cost effectiveness and lack of potential complications, are we back to square one? Treat on clinical suspicion of pneumonia, using a broad based antibiotic regimen that takes local flora into account, await the tracheal aspirate and tailor therapy?

6 Comments

Got something to say?

Log in to your account

Forgotten your password?

Todd Fraser wrote 05-20-2011 04:01:13 pm
Two very interesting, yet conflicting studies are in our library on this subject.

The Canadian Critical Care Trials Group did a study in 2006 that demonstrated no change in mortality, ventilation time, bugs not covered by empiric therapy or antibiotic usage between invasive (BAL) and non-quantitative tracheal aspirates.

This conflicts with a British study in 2008 that found overdiagnosis of VAP by tracheal aspirates (IF you use BAL as the gold standard) by a whopping 89% vrs 21%. They described reductions in antibiotic usage too.

So how do these studies mesh? A number of other studies are out there, with conflicting results.

Oliver Arkell wrote 06-08-2011 10:23:34 pm
My practice is now to start broad spectrum antibiotics based on what unit antibiogram, after taking an endotracheal aspirate / washing. We don't do quantitative analysis. We then tailor our therapy based on the results we get, and stop antibiotics if we get negative cultures.

The situation I have trouble with is when the patient is already on antibiotics - for example they've got abdo sepsis. Is it reasonable to stop antibiotics if your cultures are negative?

Jo Butler wrote 06-13-2011 10:24:28 am
Wasn't there a French paper (Fagon?) that showed a reduction in antibiotic usage when invasive methods were used? There was no increase in recurrence or other outcomes from memory - the idea being you could restrict ABs based on BAL and the patients did just as well.

Alex McKenzie wrote 06-17-2011 10:13:41 pm
This is extremely frustrating - a quick scan of the literature over the past 3 years shows an array of diagnostic criteria for VAP - almost as many as there are papers. One of the more recent I could find was the early tracheostomy paper from earlier this year, and it used quantitative BAL only. A couple of years ago the continuous subglottic suction paper in cardiac surgery used quantitative tracheal aspirates. And so on and so on.

Does anyone know of a consensus position on this?

Benjamin Moran wrote 08-19-2011 01:17:45 pm
Obviously, if you can't diagnose a disease, you can't treat it effectively. Looking at the literature, VAP is associated with a two-fold increase in mortality, increase in morbidity, length of ICU stay, and excess cost (Safdar et al, 2005). This was a meta-analysis, with the majority of diagnoses being made clinically (using the CDC definition). The incidence was between 10-20%. Looking at the studies of treatment options to aim to prevent VAP (subglottic suctioning, silver-coated ETTs), rates of VAP were reduced, but mortality rates were not significantly different. Surely, with VAP causing a 2-fold increase in mortality, if the intervention really decreased VAP, a reduction in mortality would be seen. As for selective oral and digestive decontamination, reduced rates of mortality and VAP are seen, but using systemic antimicrobials may muddy the waters somewhat. This suggests that whatever intervention reduces mortality, then the diagnostic criteria used in these studies may be where the monies at. Maybe this is a bit too simplistic...

My approach is that, given the gold-standard diagnosis of VAP eludes us, and the potential increased rate of mortality, prompt and early commencement of antimicrobial therapy should be undertaken. I use the CDC definition (new infiltrate on >=2 CXRs, fever/leucocytosis/leucopaenia, plus at least 2 of purulent secretions/tachypnoea/bronchial breath sounds/increase in O2 requirements). A sample should be taken (blind BAL is as efficacious as

Todd Fraser wrote 10-07-2012 10:35:55 am
I've just recorded a fantastic podcast interview with Rob Boots from Royal Brisbane and Women's hospital. Rob has a PhD in VAP prevention, and I'll be releasing this terrific pod in the next few weeks - stay tuned!