Previous studies investigating methods for differentiating viral and bacterial pneumonia in children, though often hampered by incomplete microbiological approaches, have found such tests as chest x-ray, white blood cell count, erythrocyte sedimentation rate, and C-reactive protein to be insufficiently reliable, the authors explain.

Dr. Olli Ruuskanen and colleagues, from Turku University Hospital in Finland, undertook a 3-year study to investigate a variety of methods of distinguishing viral and bacterial pneumonia, using several different laboratory techniques to search for 17 causative infectious agents.

Among children with bacterial pneumonia only, 72% had alveolar infiltrates on chest x-ray, the authors report, compared with 49% of children with viral pneumonia only. Alveolar infiltrates were more commonly lobar in children with bacterial pneumonia (36%) than in children with viral pneumonia (15%).

In contrast, half the children with interstitial infiltrates had viral pneumonia and half had bacterial pneumonia, the results indicate.

Elevated white blood cell counts and erythrocyte sedimentation rates were equally common in bacterial and viral pneumonia, the researchers note, but C-reactive protein levels were significantly higher among children with bacterial pneumonia. Still, a high serum C-reactive protein concentration (above 80 mg/L) significantly predicted bacterial pneumonia only in children under age 2 years.

"Because it is impossible to distinguish between viral pneumonia and bacterial pneumonia," Dr. Ruuskanen told Reuters Health, "we recommend that all children with pneumonia should be treated with antibiotics."

"In fact," Dr. Ruuskanen said, "in addition to diagnostic chest radiograph no other examinations are necessary. Measuring C-reactive protein gives only further information of the intensity of the inflammation, but it does not aid in the management."

Thorax 2002;57:438-441.