Re-evaluation of rejection criteria for endotracheal tube (ETT) specimens from adult patients.

The purpose of this study was to determine optimal criteria for microbiology laboratory screening of endotracheal tube (ETT) specimens submitted for bacterial culture from adult patients. ETT specimens from adult patients that were received by two microbiology laboratories were prospectively evaluated and subdivided into one of three study arms with the following criteria: <10 squamous epithelial cells (SECs) per low-power field with bacteria seen on Gram staining (arm 1), >10 SECs per low-power field with bacteria seen on Gram staining (arm 2) and <10 SECs per low-power field with no bacteria seen on Gram staining (arm 3). A fourth study arm (>10 SECs per low-power field with no bacteria seen on Gram staining) was planned but this arm was terminated due to the paucity of specimens meeting these criteria. Isolate evaluation was performed using standard microbiology protocols. A limited chart review was undertaken at one of the institutions, only reviewing patients from which a potential pathogen was recovered on culture. In total, 141 ETT specimens were evaluated. A potential respiratory pathogen was recovered from 54, 37 and 10 % of specimens in study arms 1, 2, and 3, respectively (P<0.0001, comparing between arm 1 and arm 3). For the 23 patients included in the chart review from whom a potential pathogen was recovered on culture, respiratory infection was considered to be present in 50 % (6/12) of patients in arm 1, 66.6 % (6/9) of patients in arm 2 and 100 % (2/2) of patients in arm 3. Therapy was rarely altered based on culture results. In this study, the ETT specimens submitted for bacterial culture were of limited benefit to clinicians. The data presented here support the use of an absence of bacteria on Gram staining as a rejection criterion for ETT specimens. The criterion of >10 SECs per low-power field should be further evaluated in a prospective study of patients with an unequivocal clinical diagnosis of pneumonia.