OBJECTIVE:

To describe and evaluate empiric antimicrobial regimens chosen for hospitalized patients with presumed community-acquired pneumonia (CAP) in US hospitals, including compliance with the American Thoracic Society (ATS) guidelines. Secondary outcomes included length of stay (LOS) and mortality associated with the choice of therapy.

METHODS:

A nonrandomized, prospective, observational study was performed in 72 nonteaching hospitals affiliated with a national group purchasing organization. Patients with an admission diagnosis of physician-presumed CAP and an X-ray taken within 72 hours of admission were eligible for the study. Demographic, antibiotic selection, and outcomes data were collected prospectively from patient charts.

RESULTS:

3035 patients were enrolled; 2963 were eligible for analysis. Compliance with the ATS guidelines was 81% in patients with nonsevere CAP. The most common antibiotic regimen used for empiric treatment was ceftriaxone alone or in combination with a macrolide (42%). The overall mortality rate was 5.5%. The addition of a macrolide to either a second- or third-generation cephalosporin or a beta-lactam/beta-lactamase inhibitor was associated with decreased mortality and reduced LOS.

CONCLUSIONS:

Most hospitalized patients with CAP receive antimicrobial therapy consistent with the ATS guidelines. The addition of a macrolide may be associated with improved patient outcomes.