Objective: To determine the techniques used for the etiological diagnosis of community-acquired pneumonia in patients admitted to the intensive care unit (ICU) and to describe the predominant causative organisms as well as prognostic factors of ICU mortality.
Patients and methods: A total of 262 patients with community-acquired pneumonia admitted to 26 ICUs between 1 November of 1991 and 31 October of 1992 were included in a prospective, open, multicenter study.
Results: The diagnostic techniques most frequently used were blood culture (243 cases) and simple tracheal aspirate (166 cases). Simple tracheal aspirate (58.4%), bronchoalveolar lavage (47.7%), and protected-specimen brush (44.2%) were the techniques that showed the highest diagnostic reliability. In 220 cases, techniques considered of high diagnostic probability were employed. With the use of these procedures, the most frequent causative pathogens were Streptococcus pneumoniae (13.6%) and Legionella pneumophila (9.5%). In 100 cases (45.5%), no pathogen was isolated. A total of 88 patients (33.6%) died during the ICU stay. Predictive variables of poor outcome selected by means of a multivariate analysis were as follows: multisystemic failure (OR = 28.6; 95% CI: 12.8-65.1; p = 0.0001), APACHE II at the time of ICU admission (OR = 5.3; 95% CI: 2.5-11.3; p = 0.0001), progression and/or spread of lung infection (OR = 4.5; 95% CI: 2.4-8.4; p = 0.0001), and shock on admission (OR = 8.48; 95% CI: 4.5-15.9; p = 0.0001).
Conclusions: In 45.5% of patients with community-acquired pneumonia admitted to ICU, no causative pathogen was identified. The prognosis of these patients was influenced by the severity of disease assessed by APACHE II score and presence of multisystemic failure and shock at the time of ICU admission.