Diagnosing Ventilator Associated Pneumonia

A 64 year old lady who had been admitted with acute pancreatitis due to gallstones. She was initially admitted to the intensive care unit for cardiovascular management and management of her electrolyte imbalance. After a few days she was intubated for hypoxia.

She developed pancreatic necrosis and pseudocyst formation. These were drained by percutaneous drains whereby she showed some improvement with more stability in her cardiovascular system. She had two failed extubations and then had a tracheostomy placed. She was weaned from the ventilator but then remained on 40-45% of oxygen for a number of weeks. Serial scans showed a static nature to her pseudocysts. Her inflammatory markers remained static at a moderate level over this time. It was felt that she had a ventilator associated pneumonia and was started on antibtiotics. She then showed improvement a number of days later. She was further weaned from the ventilator and decannulated. She needed recannulation later and suffered another episode of ventilator associated pneumonia which was treated. Eventually a number of months later she was discharged to the ward and then home.

How can we diagnose ventilator associated pneumonia?

James Day

Ventilator associated pneumonia (VAP) is common in the intensive care unit (ICU) affecting 8-20% of ICU patients and up to 27% of mechanically ventilated patients (1). The mortality rate associated with VAP range from 20- 50% and up to 70% in cases of multi-resistant organisms (2). There is increased morbidity associated with VAP including increased ICU length of stay and hospitalisation. There is a problem with delayed diagnosis and associated worse outcomes but also an incorrect diagnosis will lead to inappropriate treatment and exposure to treatment related complications. There have been many criteria used to try to diagnose VAP including clinical, imaging, microbioloigcal and bio-markers. There remains no gold standard and hence nothing to compare to when a new technique is devised.

Chest radiology is sensitive but non specific. Johanson (3) proposed using the development of consolidation on chest x-ray with 2 of the following: pyrexia greater than 38°C; raised or lowered white cell count and purulent secretions. The sensitivity was 69% with a specificity of 75% when compared to post- mortem lung biopsies. Other scoring systems have been proposed including the NNIS (National Nosocomial Infection Surveillance) system. When compared to bronchoalveolar lavage (BAL) fluid cultures the sensitivity was 84% and specificity 69%. Pugin (4) devised the Clinical Pulmonary Infection Score (CPIS) using 6 variables: fever, leukocytosis, tracheal aspirates, oxygenation, radiographic infiltrates and semi-quantitative cultures of tracheal aspirates with Gram stain. Studies have tried to validate the use of the CPIS but is limited by BAL culture not being a true gold standard.

Other studies have looked at different bacteriological sampling techniques to improve diagnosis of VAP. The different techniques include BAL; protected BAL (pBAL), protected specimen brush (PSB) and tracheobronchial aspirate (TBA). These different techniques all seem to be equivalent in the diagnosis of VAP (5). Prior antibiotic use considerably decreased the sensitivity of the cultures of BAL samples. Often the micro-organism found in quantitative culture odes not correlate with the culture obtained from pathological samples (6). Blood culture positivity in VAP ranges from 8-20%. As cultures take time to grow there has been interest in the cytological characteristics of the samples taken. These include the number of inflammatory cells and performing Gram stains. It seems that prior antibiotic use affects the usefulness of these tests and also the type of infecting organism too. Pseudomonas aeruginosa infection decreases the sensitivity of using cytological data compared to other bacteria.

It seems that bacterial sampling does not increase diagnosis accuracy compared to clinical scoring systems. The way the bacteriological samples are collected does not seem to affect accuracy. Cytological data may be useful but is limited by prior antibiotic use and the infecting organism.

Other tests that may improve accuracy include the use of biomarkers. These include C reactive protein (CRP), procalcitonin (PCT), elastin fibre (EF), soluble triggering receptor expressed in myeloid cells (sTREM) and endotoxin. Studies have shown that CRP, PCT and sTREM may aid the diagnosis of VAP but EF and endotoxin are of limited value.

The main problem concerning the performance of diagnostic tests in diagnosing VAPs is the lack of a gold standard. Using post-mortem specimens are problematic as these patients are not typical of all patients with VAP. There is also a high incidence of co-existing lung pathology in ICU patients such as thromboembolic disease, atelectasis, fibrosis and diffuse alveolar damage. Using a purely clinical approach will tend to lead to over treatment of patients. As no one diagnostic method is satisfactory it seems

prudent to incorporate several criteria. These should include imaging, bacteriological sampling and the use of biomarkers. Due to unacceptable patient morbidity and mortality incurred through delay in antibiotic prescription, it is important to start antibiotics when there is clinical suspicion of a VAP. The role of bacteriological samples comes in refining the choice of antibiotics and timing of de-escalation of antibiotic therapy.


Lessons learnt

It seems that diagnosis of VAP is difficult and there is no definitive way to be sure of your diagnosis. The decision to administer antibiotics is weighted on one side by a delay in doing so to most likely cause serious morbidity and even mortality to the patient. On the other hand unnecessary antibiotic use could lead to drug side effects and the occurrence of resistant organisms. This balance seems to be weighted towards the early and more liberal use of antibiotics if a VAP is presumed.

It seems best to use a multimodal approach using clinical scores, imaging of the chest and some kind of bacteriological surveillance sampling. The use of biomarkers seems to show promise but needs further studies to confirm. This approach should then be used to rationalise the antibiotic course either through length of course or broadness of spectrum.



  1. Tejerina E, Frutos-Vivar F, Restrepo MI, Anzueto A, Abroug F, Palizas F, González M, D’Empaire G, Apezteguía C, Esteban A, Internacional Mechanical Ventilation Study Group: Incidence, risk factors, and outcome of ventilator-associated pneumonia. J Crit Care 2006, 21:56-65.
  2. Heyland DK, Cook DJ, Griffith L, Keenan SP, Brun-Bruisson C: The attributable morbidity and mortality of ventilator associated pneumonia in the critically ill patient. The Canadian Critical Trials Group. Am J Respir Crit Care Med 1999, 159:1249-1256.
  3. Johanson WG Jr, Pierce AK, Sanford JP, Thomas GD: Nosocomial respiratory infections with gram-negative bacilli. The significance of colonization of the respiratory tract. Ann Intern Med 1972, 77:701-706.
  4. Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM: Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic ‘blind’ bronchoalveolar lavage fluid. Am Rev Respir Dis 1991, 143:1121-1129.
  5. Torres A, el-Ebiary M, Padró L, Gonzalez J, de la Bellacasa JP, Ramirez J, Xaubet A, Ferrer M, Rodriguez-Roisin R: Validation of different techniques for the diagnosis of ventilator-associated pneumonia. Comparison with immediate postmortem pulmo- nary biopsy. Am J Respir Crit Care Med 1994, 149:324-331.
  6. Kirtland SH, Corley DE, Winterbauer RH, Springmeyer SC, Casey KR, Hampson NB, Dreis DF: The diagnosis of ventilator-associated pneumonia: a comparison of histologic, microbiologic, and clinical criteria. Chest 1997, 112:445-457.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s