Severe community-acquired pneumonia: an Australian perspective

Severe community-acquired pneumonia (CAP) is a common disease with a relatively high mortality. The initial treatment is empirical, based on a broad range of potential pathogens. There are minimal published data describing microbiological causes of pneumonia in Australia.

To describe the aetiology and characteristics of severe CAP in patients requiring intensive care unit (ICU) admission, to identify factors predicting mortality and to audit current practices of investigation and antibiotic management of these patients from an Australian perspective.

A retrospective analysis of patient case notes was performed for 96 consecutive patients admitted to two ICU with severe CAP. Data recorded included patient demographics, comorbidities, antimicrobial treatment, investigations and outcome (mortality, length of stay).

Overall, mortality was 32%. A microbiological diagnosis was made in 46% of patients. The most frequent causative organisms were Streptococcus pneumoniae (13 cases), influenza A (9), Haemophilus influenzae (5) and Staphylococcus aureus (4); aerobic Gram-negative bacilli collectively accounted for five cases. Blood cultures were positive in 20% of patients. Seventy patients (73%) required mechanical ventilation and 61 patients (63%) required inotropic support. Laboratory abnormalities including acute renal failure, metabolic acidosis and coagulopathy were frequent. Factors associated with mortality on multivariate analysis were age, antibiotic administration prior to hospital presentation, delay in hospital antibiotic administration of more than 4 h, and presence of multilobar or bilateral consolidation on chest X-ray.

Severe CAP requiring ICU admission was associated with a mortality rate of 32%, despite appropriate antimicrobial therapy including a beta-lactam and a macrolide antibiotic in most cases. Causative organisms identified were similar to those found in previous studies. High rates of viral causes (28% of identified pathogens) were noted. Low rates of legionellosis and other atypical causes were found, most probably due to a lack of systematic testing for these agents.