Nosocomial pneumonia: importance of recognition of aetiological agents to define an appropriate initial empirical therapy.

Of the 16,024 patients hospitalized from January 1995 to October 1997, 397 (2.4%) acquired nosocomial pneumonia and the aetiological agent was defined in 101 (25%). About 82% developed late onset pneumonia (>7 days of hospitalization). The site of isolation of microorganisms was bronchoalveolar lavage (BAL) (49%), blood culture (39%), pleural effusion (10%) and (2%) pulmonary tissue. Gram-negative bacteria were responsible for 54% of infections. Staphylococcus aureus (34%) was the most frequent microorganism isolated followed by Acinetobacter baumannii (29%) and Pseudomonas aeruginosa and Klebsiella pneumoniae, responsible for 7% of nosocomial pneumonia. Monotherapy gave good cover for early onset pneumonia, but not for late onset pneumonia. Based on our selection criteria, ciprofloxacin was the best monotherapy for early (50%) and late (31%) onset pneumonia. Vancomycin plus ciprofloxacin gave cover of 85% of early and 64% of late onset pneumonia. Monotherapy with ciprofloxacin and also combination therapy of ciprofloxacin plus vancomycin may be good options as initial empirical therapy for nosocomial pneumonia.