To the Editor.
—We read with interest the study by Dr Bregeon and colleagues1 regarding ventilator-associated pneumonia (VAP). From their data, the authors conclude that "patient outcome is the same if VAP has been diagnosed by protected specimen brush (PSB) or by another sampling method." We feel this conclusion may be misleading.The subject of VAP is complex and still very controversial. Ventilator-associated pneumonia is common in patients receiving ventilory support; however, the impact of this complication on patient outcome is unclear. It is now generally accepted that the use of clinical diagnostic criteria alone will grossly overdiagnose VAP.2 Similarly, the specificity of tracheal aspirates is very low. Furthermore, it is important to make a microbiological diagnosis, since empiric treatment provides inadequate antimicrobial coverage for as many as 40% of patients subsequently proven to have VAP.2 Although there is no true criterion standard to diagnose VAP at