Abstract

BACKGROUND: Ten years ago we published a study of 50 adults with severe community acquired pneumonia admitted to our intensive care unit and subsequently introduced guidelines for the management of severe community acquired pneumonia which are largely in accordance with those of the British Thoracic Society. The results of a follow up study are now reported in order to assess their impact on the outcome of this disease. METHODS: Fifty seven cases of severe community acquired pneumonia admitted to our ICU between 1984 and 1993 were studied. Causal pathogens, clinical and laboratory features of severity, antibiotic therapy and mortality were studied and, where possible, compared with results from the previous study. RESULTS: Streptococcus pneumoniae, Legionella pneumophila and Staphylococcus aureus were the most frequent causes of severe community acquired pneumonia, as in the previous study. The intensity of microbial investigation has increased, particularly with regard to pneumococcal and Legionella antigen testing, the latter allowing earlier diagnosis of Legionella infection than previously. In spite of this, no pathogen was identified in 33% of cases compared with 18% previously. Indices of severity of illness were widely recognised, and a decrease in unplanned transfers to the ICU following "unexpected" cardiorespiratory arrest from 25% to 7% (p < 0.02) was found. Antibiotic therapy largely reflected guideline recommendations with 98% receiving a beta-lactam agent and 91% erythromycin. The overall mortality was 58% compared with 54% previously. CONCLUSIONS: Management guidelines for severe community acquired pneumonia have been widely adopted but without a reduction in mortality in our hospital. Factors other than early diagnosis, appropriate antibiotics, or prompt ICU transfer may influence the outcome in severe community acquired pneumonia.