Community Acquired Pneumonia



A specific etiological diagnosis is often difficult or impossible to make in community-acquired pneumonia. Given this, the true incidence of each microbial etiology is unknown. In an other­wise healthy host, probably 95 to 99 per cent of community-acquired pneumonias are due to viral, mycoplasma, pneumococcal (Streptococcus pneumoniae) or Legionella infections, and some authors estimate that viruses and mycoplasma ac­count for more than half. In patients with under­lying disease, other organisms may be found which fortunately tend to follow a certain pre­dictable pattern: Streptococcus pneumoniae, Staphylococcus aureus, and Haemophilus influ­enzae are commonly responsible for bacterial pneumonia following influenza; S. pneumoniae and H. influenzae in patients with chronic bron­chitis; and anaerobes in patients with pneumonia following aspiration of oropharyngeal contents. Highly resistant gram-negative bacilli are very rarely responsible for community-acquired pneu­monia.

If pneumonia is suspected, a physical exami­nation is essential; however, even in the absence of abnormal physical signs, such as crackles or evidence of consolidation, pneumonia may be present. A chest x-ray is required to differentiate pneumonia (infiltrate expected) from bronchitis. While certain radiological patterns have been as­sociated with different organisms, such associa­tions are of a nonspecific nature and are of little value in an individual patient. Serology may be obtained if viral, mycoplasma, or Legionella in­fection is suspected but does not aid immediate clinical decisions because a delay is incurred in demonstrating a rising titer (up to five weeks). Sputum cultures are also of limited value. Sen­sitivity is only 50 per cent for pneumococci, and other fastidious organisms are difficult to culture or are overgrown by oral flora. If positive, cultures of blood or pleural fluid provide a precise etio­logical diagnosis, but negative results are much more common. Sputum Gram’s stain remains the time-honored approach, but some patients are un­able to expectorate, and even when sputum is ob­tained, it is usually of poor quality and contam­inated by oral flora 75 per cent of the time. When an optimal specimen is obtained with extreme care, there is still a 38 per cent false-negative rate for pneumococcal pneumonia. Invasive tech­niques to obtain selective cultures are rarely re­quired in community-acquired pneumonia.

TheraPV is commonly instituted on an empir­ical basis because of the limited information pro­vided by the above diagnostic approaches. One suggested approach is depicted in Figure 23-1. In the normal host, a viral etiology is suggested by the presence of a dry cough and systemic illness jn a young person. A bacterial infection is com­moner in adult patients, who often present sud­denly with rigors, productive cough, chest pain, and a high white cell count. Patients with Legion­ella infection have features of both. Erythro­mycin is the most practical empirical agent, as it is the drug of choice for mycoplasma and Legion­ella infection and is also useful in pneumo­coccal pneumonia. The antibiotic selected for an abnormal host depends on the specific circum­stances . The decision of whether to hospitalize the patient is often based on the pa­tient’s underlying disease status.