A study of the etiologies and treatment of nosocomial pneumonia in a community-based teaching hospital.

OBJECTIVE: To compare the frequency of the pathogens of nosocomial pneumonia in a community-based teaching hospital to the frequencies previously published, and to evaluate recommendations for the therapy of nosocomial pneumonia in this setting. DESIGN: Retrospective review of prospectively acquired data accrued during 9 randomized single-blinded and 4 single-agent investigational antibiotic studies for the therapy of pneumonia in hospitalized patients between 1981 and 1989. SETTING: The study was performed at a university affiliated, community-based teaching Department of Veterans Affairs Medical Center. PATIENTS: Patients were hospitalized on the acute medical/surgical and intermediate medicine wards. Informed consent was obtained prior to enrolling patients into the respective antimicrobial studies. Pneumonia was documented radiographically and clinically for each patient. RESULTS: Two hundred thirty-one episodes of nosocomial pneumonia were treated. Overall, 51% of pneumonias were caused by Streptococcus pneumoniae or Hemophilus influenzae with or without other organisms that were not gram-negative bacilli. Gram-negative bacilli, with or without other organisms, accounted for only 26% of all nosocomial pneumonias. Overall, monotherapy with a cephalosporin (usually a broad-spectrum agent) was equally efficacious compared with combination therapy (87% versus 81%, respectively). Cure rates for nosocomial pneumonias from gram-negative bacilli treated with these 2 therapies also were similar (70% versus 60%, respectively). CONCLUSIONS: In nontertiary care settings, gram-negative bacilli may cause fewer episodes of nosocomial pneumonia (26% in this study) than noted by previously published reports, which indicated that these organisms account for 50% of nosocomial pneumonias. Further, S pneumoniae and H influenzae may account etiologically for many of these nosocomial pneumonias. Monotherapy with an extended-spectrum cephalosporin may be more appropriate than combined treatment with a beta-lactam and an aminoglycoside in a nontertiary care setting, thereby reducing potential toxicity in an older, hospitalized patient population.