SESSION TYPE: ICU Safety and Improvement Strategies
PRESENTED ON: Tuesday, October 23, 2012 at 02:45 PM - 04:15 PM
PURPOSE: In patients with suspected infectious lung pathologies, sputum samples are frequently collected prior to obtaining samples by bronchoscopic methods. However, the agreement in diagnostic yield between bronchoalveolar lavage (BAL) and sputum culture in this population is not known. To measure the agreement between these two methods, we compared the microbiologic and cytological findings of culture material obtained by BAL/wash and endotracheal tube/tracheal tube aspiration.
METHODS: We retrospectively reviewed the medical records of 270 critically ill patients who received a diagnostic unprotected BAL or bronchial wash over a four-year period. We collected data on patient demographics, co-morbidities, immune status, microbiology, cytological findings, and antimicrobial therapy. Thirty cases were eliminated due to incomplete data. Of the 240 remaining patients, 196 were intubated.
RESULTS: In 56% of patients, we observed a total agreement between the microbiology obtained by sputum culture and the sample obtained by bronchoscopy. Complete agreement between the two methods improved to 61% after controlling for contamination with Candida albicans or unspecified yeast. For individual pathogens the agreement between the two methods, after accounting for chance occurrence, was significantly higher: MSSA (k= 0.662 95% CI 0.475 --- 0.848); MRSA (k = 0.655 95% CI 0.482--- 0.828); Pseudomonas aeruginosa (k = 0.499 95% CI 0.326--- 0.673); Klebsiella pneumoniae (k= 0.583 95% CI 0.393 --- 0.772). In 25% of patients, culture material obtained by bronchoscopy grew an organism that sputum did not. Although antibiotic management changed post BAL in 85% of patients (204/240.), these changes were the result of a new finding on BAL in only 21% of cases.
CONCLUSIONS: In critically ill patients diagnosed with respiratory failure due to an infectious etiology, there was a strong agreement between microbiologic findings obtained by bronchoscopic methods and sputum culture for important pathogens. For most patients, new findings on BAL did not lead to changes in antibiotics.
CLINICAL IMPLICATIONS: In select critically ill patients with respiratory failure due to infectious etiology, sputum culture may be a more cost effective and less invasive diagnostic test.
DISCLOSURE: The following authors have nothing to disclose: Shanti Akers, James Brown, Diana Kolman, Shirish Amrutia, James Gasperino
No Product/Research Disclosure InformationDrexel University College of Medicine, Philadelphia, PA