The classic microbiological approach to community-acquired pneumonia is based on examination of easily obtainable samples (respiratory secretions, blood, or pleural fluid if available) for stains and culture. Serology has been the basic method for demonstrating the presence of atypical pathogens that are difficult to culture in non-specialized laboratories. The general effectiveness of this classic approach is far from optimal; only a minority of cases can be diagnosed during the episode, and obtaining a reliable diagnosis before antibiotic prescription is rare. Several new 'rapid' tests, generally based on immunological techniques, have been introduced to improve the yield of older methods. However, very few have satisfied a cost-benefit analysis. Molecular biology appears to be a promising alternative, but few tests are commercially available yet. Finally, more reliable respiratory samples can be obtained by invasive procedures such as bronchoscopy or direct transthoracic aspiration, but, owing to potential complications, their use is reserved for more severe cases.
The sputum Gram stain has been the cornerstone of the microbiological investigation of community-acquired pneumonia for many years. Although its diagnostic efficacy is a matter of debate, it is a simple, rapid, and economical first step in patients not previously treated with antibiotics or with sputum or tracheal aspirate microbiological processing. The predominance of a particular bacterial morphology (Gram-positive diplococci, Gram-negative bacilli, Gram-negative coccobacilli, etc.) is a good indication of the most likely pathogen. In addition, the absence of bacteria in a sample with polymorphonuclear cells suggests Legionella or an atypical agent. Culture of respiratory secretions is of less value because contaminant flora can overgrow the real pathogens. Consequently, cultures are accepted only if they agree with a Gram stain. In practice, cultures are particularly useful for demonstrating obligate pathogens such as L pneumophila or M. tuberculosis.
Pneumococcal antigen detection by counter-immunoelectrophoresis, latex agglutination, coagglutination, and enzyme immunoassay has been proposed in order to improve the sensitivity of Gram stain in pneumococcal pneumonia. However, these techniques have not been found to be cost effective, and consequently they are seldom carried out in clinical practice. Antigen detection for H. influenzae is only available for typable strains that seldom produce pneumonia in adults. Direct fluorescent antigen detection of L pneumophila in sputum may be useful in selected cases although it is insensitive. The use of molecular biology techniques in sputum samples has recently been introduced for obligate pathogens such as L pneumophila. Unfortunately, DNA probes were not very sensitive for detecting this organism.
Usually two blood cultures are routinely collected before antibiotic treatment is started. The sensitivity of this procedure is low (about 25 per cent in the case of pneumococcal pneumonia) although it is highly specific. The presence of bacteremia increases the risk of complications, and so the result of this procedure also has prognostic implications.
If there is significant pleural effusion, the pleural fluid should always be analyzed in order to exclude empyema. The diagnostic efficacy of pleural fluid Gram stain and culture is low. Detection of pneumococcal antigen in pleural fluid by latex agglutination could be a simple and effective diagnostic method when pleural effusion is present.
Detection of pneumococcal antigen in urine by counter-immunoelectrophoresis has an acceptable sensitivity. However, it is a cumbersome procedure. Detection of L pneumophila serogroup 1 antigen in urine using latex agglutination or enzyme immunoassay is commercially available and seems to have an acceptable diagnostic efficacy.
Serological testing for atypical bacteria is of little help in the acute phase of pneumonia. Paired acute and convalescent serological testing usually provides a reliable retrospective diagnosis, useful only in epidemiological studies. The detection of IgM antibodies in acute M. pneumoniae pneumonia has proved to be a useful test. Serology for viruses has little practical use.
Transtracheal needle aspiration, although still recommended in textbooks, is seldom performed owing to its poor tolerance. Although it has some lack of specificity, its sensitivity is very high. Minor complications have been reported occasionally. This procedure could be considered in non-intubated severe pneumonia without expectoration.
Among bronchoscopic methods, the protected specimen brush has been widely demonstrated to be highly sensitive and specific provided that the patient is not already on antibiotics. In this case, protected specimen brush can produce a false-negative result even after a brief antibiotic treatment (24 h). In practice, bronchoscopy and protected specimen brush are rarely carried out in non-intubated patients because of their complexity and the risk of precipitating the need for mechanical ventilation. Bronchoalveolar lavage has no advantage over protected specimen brush for diagnosing pyogenic bacterial infection; however, it is more useful for demonstrating opportunistic organisms such as Pneumocystis carini. Often, this infection presents as a severe community-acquired pneumonia.
Transthoracic needle aspiration has good sensitivity and excellent specificity. Complications are minimal when the 'ultrathin' 25G needle is used, although mechanical ventilation is considered to be an absolute contraindication. Pneumococcal antigen detection by latex agglutination in the aspirated material substantially increases the sensitivity. Despite its simplicity, safety, and effectiveness, transthoracic needle aspiration with an ultrathin needle is seldom performed in practice.
Was this article helpful?