The use of sputum as a diagnostic tool in evaluating patients with community-acquired pneumonia (CAP) is controversial. Recent studies have questioned any positive impact of sputum examination on treatment strategies. However, treatment guidelines that insist on using sputum results to direct treatment decisions in CAP persist. A recent update of the American Thoracic Society guidelines suggests collecting sputum samples only when the suspected etiology is a drug-resistant organism or one not covered by empiric treatment. Ewig and associates evaluated the use of sputum in the management of patients with CAP.

A diagnostic strategy was established for patients with CAP who presented to a 200-bed primary care hospital without microbiologic facilities in Germany. This strategy included obtaining a Gram stain of the sputum and follow-up cultures. The sputum was evaluated to determine if it was purulent or mucoid. It was also graded for the presence of leukocytes and bacteria, and the absence of epithelial cells. End points recorded included the number of patients capable of producing sputum, reasons for failure to collect sputum, validity of sputum samples, and the number of definite or probable samples. The impact of sputum sample diagnostic results on antibiotic decision-making was measured.

There were 116 patients who met the diagnostic and study criteria. Of these patients, only 36 percent were able to produce a sputum sample. Older patients and those who had prior treatment with antibiotic therapy were less likely to produce sputum. Using the study criteria, 55 percent of sputum samples were considered microscopically valid. In only 10 patients (9 percent) were diagnostic results yielded. The impact on antibiotic decision-making was minimal; antibiotics were changed on the basis of sputum results in only one case.

The authors conclude that sputum has a low diagnostic yield in the management of CAP in the primary care setting. In this study population, the number of adequate sputum samples was small, and the diagnostic yield was minimal. The authors state that sputum studies should not be part of a management algorithm in the treatment of CAP in primary care.