DIAGNOSTIC BASTERIOLOGYMicrobiology and the diagnosis and treatment of ventilator-associated pneumoniaVan Eldere, Johan JAuthor Information From the Department of Microbiology and Immunology, Catholic University of Leuven, Belgium. Correspondence to J. Van Eldere, Centrum moleculaire diagnostiek, CDG 8e verdiep, UZ Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. E-mail: [email protected] Reviews in Medical Microbiology: October 2004 - Volume 15 - Issue 4 - p 119-128 Buy Abstract Ventilator-associated pneumonia (VAP) is a major cause of mortality in intensive care unit patients. Prompt diagnosis and rapid and adequate antibiotic treatment significantly improve the prognosis. The role of clinical microbiology in diagnosis and treatment of VAP is controversial. The diagnosis of VAP is primarily based on clinical/radiological signs. Direct microbiological examination of respiratory samples can improve the accuracy of a clinical/radiological suspicion of VAP but cannot be used to withhold treatment in the presence of clinical/radiological suspicion. Nor is a positive direct microbiological examination in itself sufficient to start treatment in the absence of a clinical/radiological suspicion. Quantitative culture results may allow an adaptation of antibiotic treatment after 2–3 days, but without effect on patient outcome. Qualitative culture has limited value. The optimal sampling technique is a matter of debate. Some authors prefer sampling of the distal airways and lung parenchyma (broncho-alveolar lavage or protected specimen brush) because they find it has superior predictive value and leads to more frequent adaptation of antibiotic treatment; nevertheless, this does not result in reduced mortality or morbidity. Others prefer sampling of the trachea (tracheal aspiration or ETA) because it is non-invasive, easier to perform and less expensive. Studies show that Gram stain and quantitative culture of ETA offer equal if not better sensitivity than distal airway samples. Patient outcome with a diagnostic strategy based on ETA is not different from one based on distal airway sampling. Most authors consider regular surveillance cultures useful for defining an adequate empiric treatment. The predictive value of surveillance cultures for individual patients is limited and no data exist on its impact on patient outcome. © 2004 Lippincott Williams & Wilkins, Inc.