Procalcitonin levels predict bacteremia in patients with community-acquired pneumonia: a prospective cohort trial.

BACKGROUND Guidelines recommend blood culture sampling from hospitalized patients with suspected community-acquired pneumonia (CAP). However, the yield of true-positive results is low. We investigated the benefit of procalcitonin (PCT) on hospital admission to predict blood culture positivity in CAP. METHODS This was a prospective cohort study with a derivation and validation set including 925 patients with CAP who underwent blood culture sampling on hospital admission. RESULTS A total of 73 (7.9%) patients had true bacteremia (43 of 463 in the derivation cohort, 30 of 462 in the validation cohort). The area under the receiver operating characteristics curve of PCT in the derivation and validation cohorts was similar (derivation cohort, 0.83; 95% CI, 0.78-0.89; validation cohort, 0.79; 95% CI, 0.72-0.88). Overall, PCT was a significantly better predictor for blood culture positivity than WBC count, C-reactive protein, and other clinical parameters. In multivariate regression analysis, only antibiotic pretreatment (adjusted odds ratio, 0.25; P < .05) and PCT serum levels (adjusted odds ratio, 3.72; P < .001) were independent predictors. Overall, a PCT cutoff of 0.1 microg/L would enable reduction of the total number of blood cultures by 12.6% and still identify 99% of the positive blood cultures. Similarly, 0.25 microg/L and 0.5 microg/L cutoffs would enable reduction of blood cultures by 37% and 52%, respectively, and still identify 96% and 88%, respectively, of positive blood cultures. CONCLUSIONS Initial PCT level accurately predicted blood culture positivity in patients with CAP. PCT measurement has the potential to reduce the number of drawn blood cultures in the emergency department and to implement a more targeted allocation of limited health-care resources.