Diagnostic approaches and hospital-acquired pneumonia

Hospital-acquired pneumonia (HAP) is the second most frequent nosocomial infection with an average crude mortality around 40%. Clinical and microbiological diagnoses are important in order to establish an adequate antibiotic treatment. In mechanically ventilated patients, clinical diagnosis frequently leads to false-positive and false-negative interpretations, because entities other than pneumonia may cause fever and pulmonary infiltrates. Microbiological diagnostic methods are divided into invasive and noninvasive approaches. Among the former, the protected specimen brush and bronchoalveolar lavage via fiberoptic bronchoscopy are the most popular. Sensitivities and specificities range from 60 to 100%. False-negative results are mainly due to prior antibiotic treatment while false-positive results are due to distal airway colonization. Blind methods through the endotracheal tube have been as accurate as guided methods due to the anatomical distribution of ventilator-associated pneumonia (VAP) (diffuse, bilateral, and predominantly affecting the dependent lung zones). Transthoracic needle aspiration has been used by some groups in nonventilated patients. Among the noninvasive methods, quantitative culturing of endotracheal aspirates seems to offer reasonable results. Different gold standards have been used to validate all these diagnostic methods which makes comparisons very difficult. The only reliable gold standard is the presence of pneumonia in the histopathological examination of the lung. There is no clear reason to initially perform invasive testing in nonventilated patients. This is more controversial in VAP, with arguments in favor and against. However, recent information suggests that using invasive procedures does not modify the morbidity and mortality of patients with VAP but leads to a greater cost. Our personal recommendation is to start empirical antibiotic treatment according to standardized guidelines and adjust it according to quantitative cultures of endotracheal aspirates. However, thresholds of quantitative cultures (of endotracheal aspirates or any other technique) have to be flexible and balanced by clinical judgment. In those cases (ventilated and nonventilated) not-responding to initial treatment, invasive techniques may be warranted.