A comparative study of bacteremic and non-bacteremic pneumococcal pneumonia

BackgroundFew attempts have been made to compare bacteremic and non-bacteremic pneumococcal pneumonia, mainly because it is difficult to gain agreement on which cases represent non-bacteremic pneumococcal pneumonia. Recently, an immunochromatographic assay for the detection of Streptococcus pneumoniae urinary antigen has been successfully evaluated for the diagnosis of pneumococcal pneumonia. The aim of our study was to examine and compare clinical and radiological features, risk factors, and outcome associated with bacteremic and non-bacteremic groups.MethodsA retrospective study (1995–2003) analyzing the clinical records of patients diagnosed with pneumococcal pneumonia in our institution was performed. S. pneumoniae were identified by blood cultures (bacteremic group) and detection of urinary antigen (non-bacteremic group).ResultsThere were 82 patients (57 bacteremic and 25 non-bacteremic). In seven non-bacteremic cases, another etiology was detected, i.e., Legionella (n = 1) and Chlamydia pneumoniae (n = 6). Bacteremic patients were significantly younger (p = < 0.001), more likely to have liver disease (p = 0.028), current smokers (p = 0.024), alcohol and intravenous drug abusers (p = 0.014 and p < 0.001, respectively), and infected with HIV (p < 0.001). Non-bacteremic patients were more likely to have congestive heart failure (p = 0.004), chronic obstructive pulmonary disease (p = 0.033) and to be former smokers (p = 0.004). Bacteremic cases needed more prolonged intravenous antibiotic treatment (6 days vs. 4.5 days; p = 0.006) than non-bacteremic cases and their length of stay was also longer.ConclusionIn our study, smoking was the leading risk factor for pneumococcal pneumonia. However, current smokers have an increased risk of bacteremic forms and former smokers and patients with COPD developed non-bacteremic forms more frequently. Bacteremic patients need more prolonged intravenous antibiotic treatment than non-bacteremic patients.