In the Literature

Patients who receive mechanical ventilation frequently develop an increase in the volume of purulent respiratory secretions, often in association with fever and peripheral leukocytosis, despite an absence of radiographic evidence of pneumonia. These findings are consistent with the presence of acute tracheobronchitis. Although cases may occasionally have nonbacterial etiologies, such as herpes simplex virus, most are likely bacterial in origin. This complication may represent an intermediate state in a continuum that begins with airway colonization and progresses to development of VAT and then to ventilator-associated pneumonia (VAP)—although the latter may occur without intervening clinical evidence of tracheobronchitis. Consistent with the belief that tracheobronchitis serves as a way station on the path to pneumonia is the observation that a significant increase in bacterial density in bronchoalveolar lavage specimens is evident 2 days before evidence of VAP [1]. In addition, it can be hypothesized that, even if VAP is already present, concomitant airway infection may contribute to adverse outcomes and that direct application of antibiotics to the airways may improve outcomes, over and above the benefits associated with systemically administered antibiotics. Palmer and colleagues randomized adults with VAT who had undergone mechanical ventilation for a minimum of 3 days to receive either antibiotics or saline placebo by aerosolization for a planned 14 days. The diagnosis of VAT required production of 12 mL of respiratory secretions over a 4-h period, as well as bacteria visualized on a Gram-stained smear of the secretions. Vancomycin (120 mg) and/or gentamicin (80 mg) were administered every 8 h, depending on the Gram stain results for persons randomized to receive aerosolized antibiotics. Most patients (32 of 43) already had VAP and were already receiving systemic antibiotics at the time of randomization. The administration of aerosolized antibiotics was associated with statistically significant improvements in indices of respiratory infection, using both the Centers for Disease Control and Prevention–National Nosocomial Infectious Surveillance System criteria for diagnosis of VAP and the clinical pulmonary infection score, as well as improvements in the WBC count and in the frequency of newly administered systemic antibiotic therapy. The bacterial density in tracheal aspirate specimens was significantly reduced, and antibiotic resistance emerged less frequently. Weaning from mechanical ventilation was more frequently successful for persons who received aerosolized antibiotic therapy. Interpretation of the results of this study is complicated by both the coexistence of VAP and the systemic antibiotic administration in most patients. Sorting out the therapeutic and prophylactic effects of aerosol and systemic antibiotics is difficult. A meta-analysis of 8 comparative trials concluded that prophylactic administration of antibiotics directly into the respiratory tract was associated with a significant reduction in the incidence of pneumonia acquired in the intensive care unit [2]. On the other hand, a small study that examined administration of systemic antibiotics to patients with VAT, none of whom had VAP at entry, was discontinued prematurely because of apparent excess mortality among patients who had been randomized to receive no antibiotic therapy [3]. Systemic antibiotic therapy was associated with other favorable outcomes, including a reduction in the number of subsequent cases of VAP, but the study suffered from numerous shortcomings, as outlined by Craven [4] in an accompanying editorial commentary. Nonetheless, the approach described here appears to be promising, but it requires confirmation in larger studies. Such studies will require an unequivocally validated definition of VAT and will need better and clearer differentiation between VAT and VAP. This includes evaluation of the possibility that some patients with normal chest radiograph findings have pulmonary infiltrates that could be detected by CT. The etiology of infiltrates must be clearly defined, both at study entry and in persons who develop such changes during the study. Such study would be difficult, but the results would provide information critical to case management.