Severe community-acquired pneumonia - Risk factors and follow-up epidemiology

The aim of the study was to determine risk factors for severe community-acquired pneumonia (CAP) as well as to compare microbial patterns of severe CAP to a previous study from our respiratory intensive care unit (ICU) originating from 1984 to 1987. Patients admitted to the ICU according to clinical judgment were defined as having severe CAP. For the study of risk factors, a hospital-based case-control design was used, matching each patient with severe CAP to a patient hospitalized with CAP but not requiring ICU admission. Microbial investigation included noninvasive and invasive techniques. Overall, 89 patients with severe CAP were successfully matched to a control patient. The presence of an alcohol ingestion of greater than or equal to 80 g/d (odds ratio [OR] 3.9, 95% confidence interval [CI] 1.4 to 10.6, p = 0.008) was found to be an independent risk factor for severe CAP and prior ambulatory antimicrobial treatment (OR 0.37, 95% CI 0.17 to 0.79, p = 0.009) to be protective. Streptococcus pneumoniae (24%) continued to be the most frequent pathogen; however, 48% of strains were drug-resistant. "Atypical" bacterial pathogens were significantly more common (17% versus 6%, p = 0.006) and Legionella spp. less common (2% versus 14%, p = 0.004) than in our previous study, whereas gramnegative enteric bacilli (GNEB) and Pseudomonas aeruginosa continued to represent important pathogens (6% and 5%, respectively). Our findings provide additional evidence for the importance of the initiation of early empiric antimicrobial treatment for a favorable outcome of CAP. Variations of microbial patterns are only in part due to different epidemiological settings. Therefore, initial empiric antimicrobiaI treatment will also have to take into account local trends of changing microbial patterns.