Multicenter study of hospital-acquired pneumonia in non-ICU patients

Study objective: To know the incidence, epidemiology, etiology, and outcome of hospital-acquired pneumonia (HAP) in non-ICUs adult patients. Setting: Twelve Spanish teaching hospitals. Interventions: From April 1999 to November 2000, non-ICU HAP was prospectively studied by active, bimonthly 1-week surveillance. Epidemiologic data, etiology, and evolution of pneumonia were recorded. Blood and sputum cultures and Legionella pneumophila and Streptococcus pneumoniae urinary antigen tests were performed. Results: We included 186 patients, with complete data available in 165 patients (70.3% male gender; mean age, 63.7 +/- 16.9 years [+/- SD]) The mean incidence of HAP was 3 +/- 1.4 cases/1,000 hospital admissions. Most patients (64.2%) were in medical wards, had severe underlying diseases (66.6%), and had a hospital stay > 5 days (76.4%). Blood cultures were performed in 139 patients (84.2%), sputum cultures were performed in 89 patients (53.9%), and urinary antigen detection was performed in 123 patients (74.5%). An etiologic diagnosis was obtained in 60 cases (36.4%), and 31 were definitive. The most frequent etiologies were S pneumoniae (16 cases, 14 definitive), L pneumophila (7 cases, 7 definitive), Aspergillus sp (7 cases, 3 definitive), Pseudomonas aeruginosa (7 cases, 2 definitive), and several Enterobacteriaceae (8 cases, 4 definitive). Clinical complications occurred in 52.1% of the cases, and mortality was 26% (13.9% attributed to pneumonia). Conclusions: Non-ICU HAP is an important cause of hospital morbidity, observed most frequently in medical wards and elderly patients with severe underlying diseases. In this setting, S pneumoniae and Legionella sp should be considered in addition to other nosocomial pathogens; urinary antigen detection is useful in determining the prevalence of these microorganisms.