Last update:
03-Dec-2009

OBJECTIVE Blood culture (BC) is considered part of the initial laboratory examination of hospitalized patients with community-acquired pneumonia (CAP), although recent studies have challenged the necessity of such practice. A retrospective exploration was made of the contribution of blood culture to the management of CAP, the correlation of bacteremia with disease severity, and the predicted bacteremia risk.

METHOD The records were examined of 268 patients who were admitted with CAP and had blood drawn for culture (aerobic-anaerobic x2) before antibiotics were given.

RESULTS A total of 564 blood samples were cultured. Forty-one patients (15.3%) had at least one positive blood culture, of which 9 (3.4%) were true bacteremia. A blood culture isolate not covered by the empirical antibiotic therapy was identified in 3 cases (1.1%), while in 6 cases the antibiotic spectrum was narrowed following the blood culture results. All 9 significant cases of bacteremia occurred in patients aged >65 years (P=0.01) and 7 of the 9 occurred among severely diseased patients (PSI classes IV and V; P=0.03). For the comorbidities examined (diabetes mellitus, congestive heart failure, chronic renal failure, chronic liver disease and neoplasia) no statistically significant difference in bacteremia prevalence was found, apparently due to the small sample size. There were no significant differences in the incidence of bacteremia according to the decision support model for predicting bacteremia.

CONCLUSIONS Blood culture contributes to the management of CAP in a small number of patients with identifiable characteristics and increased risk for adverse outcome, but its indiscriminate use during the work-up of CAP cannot be advocated.

Key words: Bacteremia, Blood culture, Community-acquired pneumonia, Treatment.