New insights in determining the aetiology of severe community-acquired pneumonia and its implications

Abstract number: P843

Poblet-Mas N., De la Torre M., Sirvent J.M., Garcia-Gil L.J.

Objectives: Severe community-acquired pneumonia (SCAP) can be caused by different bacteria which show similar clinical presentations. Legionella pneumophila (LP) can be difficult to diagnose and its prevalence might be underestimated. The main goals of this work were:

1To evaluate the aetiology of SCAP, the prevalence of the aetiologic agents and its relationship with the clinical manifestations and the risk factors.

2To assess the use of a new PCR method specific for LP as an analysis of routine.

Methods: A total of 256 specimens (BAL, blood and urine) from 96 patients (66 with SCAP and 30 controls) admitted in the Intensive Care Unit (ICU) were analysed, from December 2004 to November 2007. Etiological agents were assessed simultaneously by culturing, by urinary antigen test (UAT) and by PCR for LP. Kendall's Tau-b correlation, Mann-Whitney and Kruskall-Wallis tests, and Fisher's Chi-square were performed, using as dependent variables SCAP patients or LP-positives by PCR. Thirty-one independent variables were analysed, twenty-three describing symptoms, seven for risk factors and one for the causative agent. The level of significance was set at P<0.05.

Results: The most prevalent pathogen was LP (45.5%), followed by Streptococcus pneumoniae (25.8%) and others (8.2%). In 20.5% of the patients no aetiologic agents were detected. Polymicrobial infections were found in 24.2% of the patients with SCAP. LP-positives obtained by PCR were significantly (P<0.001) higher than those obtained by UAT. Smoking (P<0.05) and respiratory diseases (P<0.01) were identified as risk factors. On the other hand, fever, cough, fatigue, dyspnea, sputum, tachypnea, pulmonary shunt and infiltrate, amongst others, were symptoms of SCAP (P<0.005). Significant differences between SCAP and control patients (P<0.05) in the ICU length of stay were only observed when the infection was polymicrobial. Although a significant correlation (P = 0.001) was found between SCAP and LP positives by PCR, no correlation was observed (P>0.05) between SCAP and LP positives by UAT.


1The significant symptoms and risk factors found in this study should be taken into account when diagnosing and selecting the initial antibiotic therapy.

2Polymicrobial infections present a worse prognosis and more agressive clinical evolution.

3Urinary antigen test is not sensitive enough to determine SCAP positives, while PCR is a better and more powerful tool.

Session Details

Date: 19/04/2008
Time: 00:00-00:00
Session name: 18th European Congress of Clinical Microbiology and Infectious Diseases
Location: Barcelona, Spain
Presentation type:
Back to top