Atypical Coverage Important in Community-Acquired Pneumonia

LOUISVILLE, Ky., May 15 -- Hospitalized patients with community-acquired pneumonia should get drugs that cover both typical and atypical pathogens, international researchers have recommended.

Such treatment results in a quicker time to clinical stability, shorter hospital stays, and a lower death rate, concluded Forest Arnold, D.O., of the University of Louisville, and colleagues worldwide, in the second May issue of the American Journal of Respiratory and Critical Care Medicine.

A global analysis of the occurrence rate and treatment of atypical pneumonia supports the notion that "all hospitalized patients with community-acquired pneumonia should receive empiric therapy with a regimen that covers typical and atypical pathogens," Dr. Arnold and colleagues wrote.

Atypical pneumonia is typically caused by pathogens such as Legionella pneumophila, Mycoplasma pneumoniae, and Chlamydophila pneumoniae that do not respond to ?-lactam antibiotics.

Guidelines in the U.S., Canada, Germany, Japan, and parts of Latin America urge empiric treatment for community-acquired pneumonia, with both a ?-lactam for typical pathogens and other drugs, such as macrolides or fluoroquinolones, to target the atypical pathogens.

The researchers divided the world into four regions -- North America, Europe, Latin America, and Asia and Africa.

Using an international database of 4,337 patients, kept at the University of Louisville, they calculated that the occurrence rate of atypical pneumonia was between 20% and 28%, depending on the region, from September, 1996 until April, 2004.

Specifically, the rate was 22% in North America, 28% in Europe, 21% in Latin America, and 20% in Asia and Africa.

But an analysis of treatment, using information on 2,208 patients in the Community-Acquired Pneumonia Organization database, showed marked differences in the proportion of patients who were treated for atypical pathogens.

Specifically, in North America, 91% got atypical coverage, compared with 74% in Europe, 53% in Latin America, and 10% in Asia and Africa.

The differences are important, Dr. Arnold and colleagues said, because atypical coverage can make significant differences in outcomes. Specifically:

• Patients given atypical coverage were clinically stable, on average, in 3.2 days, compared to 3.7, a difference that was significant at P<0.001.
• They stayed in hospital, on average, 6.1 days, compared to 7.1, which was significant at P<0.01.
• And the mortality rate was 7% compared to 11.1%, which was significant at P<0.05.

The American Thoracic Society defines clinical stability in terms of improved clinical signs (such as cough and shortness of breath), no fever for at least eight hours, a decrease of 10% or more in the number of white blood cells, and the ability to take oral nourishment.

The researchers pointed out that the study only applies to patients who need hospital care. In ambulatory patients, they said, "the beneficial effect of antibiotics using atypical coverage is more difficult to recognize because the time to clinical stability in this population is not measured and mortality is a very rare outcome."

The issue has been controversial, Dr. Arnold and colleagues said, because previous studies have been unclear on whether atypical coverage reduces mortality.

One reason for that lack of clarity, they said, may be that much of the mortality in hospitalized patients with CAP is not related to the pulmonary infection, so that the choice of initial therapy appears to have little effect.

The main limitation of the study, they said, is that it is retrospective, so that unmeasured confounding factors may be in play.

In addition, they pointed out that "a major limitation of the etiologic findings is that the samples came from patients enrolled in drug trials. These trials used enrollment criteria that limit the generalizability of the etiologic data presented in this study. Because this was not a randomized trial, an important limitation of the outcomes analysis is the possibility of indication bias."