First, we share the concerns of Dr. Shepherd about diagnosing ventilator-associated pneumonia (VAP). 1However, neither one of our commentaries bring a novel and accurate diagnostic approach to VAP. Dr. Shepherd was concerned whether we overdiagnosed or overtreated our patients. In the meantime, we are concerned whether the recommended invasive (bronchoscopic) diagnostic techniques 2worsen the ventilatory status and management of patients by prolonging the duration of mechanical ventilation. Both of the clinical concerns are extremely important because VAP is the leading nosocomial infection in the intensive care unit, 3and it surely increases mortality. 4
In response to Dr. Shepherd’s concern, I would like to refer to two recent studies in which the investigators tested the sensitivity and specificity of various diagnostic techniques by the help of “gold standard” histologic and microbiologic references. 5,6Torres et al. 6showed that the sensitivity (43–83%) and specificity (67–91%) ranges of both invasive and noninvasive diagnostic sampling techniques were unsatisfactory, and causative organisms were missed in a significant number (17–83%) by all techniques. Their conclusion was that all sampling techniques for detecting VAP were of limited value.
Fabregas et al. 5showed that clinical criteria—which were nearly the same as those we used—had reasonable diagnostic value. Noninvasive and invasive sampling techniques had diagnostic values comparable to clinical criteria. They concluded that an algorithm guiding antibiotic treatment exclusively by microbiologic results did not increase the overall diagnostic accuracy and would have the risk of undertreatment.
In short, diagnosis of VAP is extremely important in care of the critically ill. Moreover, we still do not have the accurate diagnostic tools.