To the Editor—Ventilator-associated pneumonia (VAP) is one of the most serious healthcare-associated infections; it has a high mortality rate, especially in intensive care units (ICUs).Reference Kalil, Metersky and Klompas ¹

Diagnosing VAP is a complex issue, and the precise role of microbiologic parameters such as cultures (if qualitative or quantitative), as well as Gram stain, remain unclear.Reference Kalil, Metersky and Klompas ^{1 ,} Reference Albert, Friedrich and Adhikari ² Microscopic evaluation by Gram stain of easily obtained respiratory secretions, such as endotracheal aspirate (EA), could provide a potentially useful guide to appropriate antimicrobial therapy in patients with suspected VAP.Reference Albert, Friedrich and Adhikari ^{2 ,} Reference O’Horo, Thompson and Safdar ³

To evaluate the performance of microscopic examinations by Gram staining of endotracheal aspirates (EAs) recovered from intensive care patients, a prospective study was performed.

Endotracheal aspirates were consecutively recovered from mechanically ventilated patients in an adult ICU in a tertiary hospital of Porto Alegre, southern Brazil, between January 1 and October 3, 2016. Smears were stained with Gram stain and were then cultured quantitatively. As selection criteria, only specimens with <10 squamous epithelial cells by microscopic examination in a low-power field were included in the study. Also, microorganisms presenting growth ≥10⁶ colony forming units (CFU) per milliliter of sample plated were considered a positive culture. Results from microscopy and culture were obtained independently and carried out by double-blind analysis.

A total of 717 EAs were obtained. Among them, 52 EAs were excluded due to the presence of >10 epithelial cells, and 13 (1.8%) were excluded due to inconsistent results in the culture (growth of non-pathogenic organisms such as yeasts). In the remaining 652 samples, a negative culture (ie, no bacterial growth ≥10⁶ CFU/mL) was observed in 415 (63.6%). Among the 237 positive cultures, gram-negative rods were recovered from 218 (92%) and gram-positive cocci were recovered from 19 (8%). For gram-positive cocci, only S. aureus were recovered in sufficient numbers to meet study criteria; only 2 of these (10.5%) were methicillin-resistant (MRSA).

Overall, 560 of 652 (85.9%) samples showed agreement between Gram stain and culture results. The sensitivity, specificity, positive and negative predictive values, and their confidence intervals (95% CI) for clustered gram-positive cocci and gram-negative rods are shown in Table 1.

TABLE 1

Sensitivity, Specificity, Positive and Negative Predictive Values of Gram-Stain Findings in Relation to the Quantitative Cultures of 652 Endotracheal Aspirates

Fast and accurate microbiological diagnosis of VAP is a major challenge, and no generally accepted gold standard exists for its diagnosis. In recent guidelines by the Infectious Diseases Society of America and the American Thoracic Society on the management of adults with hospital-acquired pneumonia and VAP, noninvasive sampling with semiquantitative cultures has been suggested instead of invasive sampling with quantitative cultures.Reference Kalil, Metersky and Klompas ¹

Although universally accepted as a useful tool for evaluating clinical specimens, the real value of the Gram stain to guide an empirical approach is also controversial. Detection of gram-positive cocci in clusters on direct microscopic examination of EAs would constitute an important tool in antimicrobial stewardship and the use of anti–gram-positive agents, especially when S. aureus is recovered. On the other hand, several studies have pointed to the low sensitivity and positive predictive values of the Gram stain, contradicting its use as a presumptive guide to therapy.Reference Albert, Friedrich and Adhikari ^{2 –} Reference Raghavendran, Wang and Belber ⁵

At our institution, a prior study showed a very high predictive negative value of the Gram stain of EAs when gram-positive cocci in clusters were considered;Reference Victorino, Dias and Ribas ⁶ this study confirmed a virtually 100% negative predictive value (Table 1). To avoid inappropriate and empirical use of vancomycin, it is important to know when not to use this drug especially in a setting with very low MRSA prevalence, such as ours. For this purpose, the Gram stain serves a crucial purpose, particularly when gram-positive cocci in clusters are concerned.

In conclusion, Gram staining of EAs showed a very high negative predictive value in this study, contributing to a more conservative use of antimicrobials. In healthcare institutions with a low VAP prevalence due to MRSA, a Gram stain of EAs without the presence of gram-positive cocci may be the strongest reason to avoid the use of vancomycin.