Objective: To evaluate the use of enriched thioglycolate as a transport medium of protected specimen brush samples, in particular its value in direct examination of respiratory specimens and in guiding initial antibiotic prescription.
Design: Prospective, randomized, pilot study.
Setting: Medical-surgical teaching intensive care unit.
Subjects: Thirty adults with suspected ventilator-associated pneumonia.
Intervention: Bronchoscopy was performed by using standard techniques, and two consecutive protected specimen brush samples were taken. Transport medium consisted of 1 mL of sterile saline or thioglycolate. Gram stains were performed in all samples. Randomization was used to select which transport medium was used first. Each patient served as his or her own control. The laboratory was blind to the choice.
Measurements and main results: Causative agents were cultured in 16 episodes, and no significant differences were observed when the transport media were compared. Anaerobes were identified in only one episode. Direct staining in thioglycolate samples anticipated the presumptive diagnosis (in presence of false-negative cultures) in three additional patients in whom prior antibiotic therapy was prescribed; this was also the case in one patient with saline solution. The etiology was anticipated by direct Gram staining and permitted a more targeted initial empirical treatment in 75% of samples transported in thioglycolate, compared with only 37.5% of samples transported in saline solution (p <.05).
Conclusion: When protected specimen brush samples are obtained, thioglycolate may contribute to early identification of the pathogen and may guide the initial empirical therapy.