Active surveillance cultures of methicillin-resistant Staphylococcus aureus as a tool to predict methicillin-resistant S. aureus ventilator-associated pneumonia

Objective: Ventilator-associated pneumonia is one of the most common infections in the intensive care unit and methicillin-resistant Staphylococcus aureus has emerged as a common cause of ventilator-associated pneumonia. We sought to study the performance characteristics of once weekly active surveillance culture of methicillin-resistant S. aureus colonization in predicting the development of methicillin-resistant S. aureus ventilator-associated pneumonia. Design: Prospective observational study. Setting: Eighty-nine-bed surgical and medical intensive care units in a university-affiliated urban teaching hospital and level I trauma and burn center. Patients: All patients >= 16 yrs old admitted to the intensive care unit on mechanical ventilation >= 48 hrs who met diagnostic criteria for ventilator-associated pneumonia by quantitative lower respiratory tract cultures obtained through bronchoscopic alveolar lavage or brush specimen between January 2008 and October 2010 were included. Interventions: None. Measurements and Main Results: Nine hundred twenty-four episodes of suspected ventilator-associated pneumonia were evaluated, and 388 patients with bronchoalveolar lavage-confirmed ventilator-associated pneumonia were included. Surveillance cultures were taken from the nares, oropharynx or trachea, and any open wound routinely on admission to the intensive care unit, every 7 days afterward, and at intensive care unit discharge. Of the 388 patients, 37 (9.5%) had methicillin-resistant S. aureus ventilator-associated pneumonia and 54 (13.9%) had methicillin-resistant S. aureus colonization documented by active surveillance culture before the development of ventilator-associated pneumonia. The sensitivity and specificity of prior methicillin-resistant S. aureus colonization as a predictor for methicillin-resistant S. aureus ventilator-associated pneumonia are 70.3% (95% confidence interval [CI] 52.8-83.6) and 92.0% (95% CI 88.5-94.5), respectively. The positive and negative predictive values are 48.1% (95% CI 34.5-62.0) and 96.7% (95% CI 94.0-98.3). Conclusions: In our study, prior methicillin-resistant S. aureus colonization as ascertained by once-weekly active surveillance culture yielded high specificity and negative predictive value, suggesting that negative active surveillance culture can accurately exclude methicillin-resistant S. aureus as an etiology in most patients with ventilator-associated pneumonia and may decrease the need for empirical methicillin-resistant S. aureus coverage in patients with suspected ventilator-associated pneumonia. (Crit Care Med 2012; 40: 1437-1442)