RESUMO
BACKGROUND: Few studies have investigated the incidence of pulmonary anaerobes in a specific population in surgical Intensive Care Unit (ICU). The objective of this work was to determine the incidence of anaerobes in surgical ICU patients with suspected pneumonia. METHODS: This was a prospective observational, single-center study. Analysis was based on data collected over 30 months from the surgical ICU of a tertiary care hospital (Rouen University Hospital), including data on risk factors for anaerobes in the lungs. Patients with suspected pneumonia (community-acquired or nosocomial) were included. Bacteriological sampling was performed by protected distal bronchial sampling (PDBS) with minilavage under bronchoscopy. Aerobic and anaerobic cultures were performed for each sample. Clinicians were only aware of aerobic results. Univariate and multivariate statistical analysis compared groups with and without anaerobes. RESULTS: A total of 134 samples were obtained from 117 patients. Surgery was performed on 74 patients (63.2%), within 24 hours of admission. Fifty-four patients (46.2%) had a chest trauma and 20 patients (17.1%) were admitted for a digestive pathology. Average age was 53.6±20.9 years and sex ratio was 5.9 (100 men/17 women). Average SAPS II was 41.6±15.1, median length of ICU stay was 23 days (25th percentile=13, 75th percentile=33), and median duration of mechanical ventilation was 21 days (25th percentile=11, 75th percentile=28). Mortality rate in ICU was 14.5%. After sampling, diagnosis of pneumonia was confirmed in 70 cases (52.2%). Anaerobe cultures were positive in 11 samples taken from 11 different patients (overall incidence 8.2%). Aerobic bacteria were also involved in 9 patients (81.8%). In univariate analysis, enteral feeding (P=0.02) and absence of catecholamines at time of sampling (P=0.003) were significantly associated with the presence of anaerobes in PDBS. Enteral nutrition was also found to be a risk factor in multivariate analysis (OR=11.8, 95% CI [1.36 to 102.4] P=0.025). Prior antianaerobic antibiotic treatment was not a protective factor. No difference was observed regarding the notion of aspiration, survival, total length of stay and duration of mechanical ventilation, or evolution of pneumonia between the two groups. CONCLUSION: Our study demonstrates the presence of anaerobic bacteria in the lung samples of patients from surgical ICU with an incidence comparable to that found in populations from medical ICU departments. Anaerobic morbidity in our study is in line with recent literature.