Tracheal aspirate Gram stain has limited sensitivity and specificity for detecting Staphylococcus aureus

Background and objective: The increasing incidence of respiratory infections due to methicillin resistant Staphylococcus aureus has resulted in increased empirical use of antibiotics active against this pathogen. There are limited data available as to whether the Gram stain of respiratory tract secretions accurately predicts growth of S. aureus. We theorized that the distinctive morphology of S. aureus would allow rapid, accurate identification of the organism in respiratory secretions. Methods: The authors reviewed all available Gram stains of tracheal aspirates sent to our hospital's microbiology laboratory between 1 April 2008 and 31 October 2008, while blinded to the culture result, and recorded the presence or absence of organisms with a morphology consistent with S. aureus. These results were correlated with the semiquantitative culture result. Results: Among 136 tracheal aspirates studied, 50 (37%) grew S. aureus. The Gram stain was read as positive for organisms consistent with S. aureus in 34 of these. Among 86 samples that did not grow S. aureus, the Gram stain was read as negative in 62. Therefore, the Gram stain had a sensitivity of 68%, a specificity of 72%, a negative predictive value of 80% and a positive predictive value of 59% for culture of S. aureus. False negative Gram stains were more likely when the culture revealed only rare or small growth of S. aureus (P = 0.01). Conclusions: In this study, the tracheal aspirate Gram stain read by an experienced clinician who was not a microbiologist, was not accurate enough to reliably predict the growth of S. aureus.