SAN FRANCISCO — Physicians who forgo obtaining cultures from patients who come in with possible community-acquired bacterial pneumonia are probably practicing wisely, John G. Bartlett, M.D., said at the annual meeting of the American College of Physicians.
It has been reported that the etiologic agent in pneumonia is never identified in 50% of cases. But that figure comes from clinical trials, in which patients are tested and cultured exhaustively, said Dr. Bartlett, chief of the division of infectious diseases at Johns Hopkins University, Baltimore.
In the hospital, the etiologic organism is identified in only 15%–20% of pneumonia cases, and the most of those results come from blood culture, not sputum.
“We don't do very well with cultures, and therefore, you can't rely on them,” he said.
Sputum rarely yields a positive culture, even in a patient with a pneumonia caused by Streptococcus pneumoniae. And blood cultures may give misleading results because they are so often contaminated, Medicare data have suggested. “There have been several reports that have shown that blood cultures really don't affect outcome in any meaningful way,” Dr. Bartlett said.
Instead, current guidelines say when diagnosing suspected pneumonia from bronchitis, an x-ray is key, although not needed in the patient with normal vital signs and no rales. If the x-ray shows an infiltrate, then the patient has pneumonia, and antibiotic treatment can be initiated empirically, with no need for a culture, because experience suggests that most patients get better with empiric treatment. “The x-ray showing an infiltrate is really a pivotal part of the diagnostic evaluation,” he said. “It separates, to a large extent, antibiotics vs. no antibiotics.”
The exception to this empiric-treatment rule is a patient who is ill enough to be hospitalized, he said. In those patients, he would order a urinary antigen test for Legionella species, because of its virulence and because it is a public health problem, and perhaps a urinary antigen test for pneumococcus, to pick up bacteremia.