BACKGROUND: Several practice guidelines for the empirical antimicrobial treatment of hospital-acquired pneumonia (HAP) have been developed, but the acceptance and use of such guidelines are unknown.
OBJECTIVE: To assess physicians' use of empirical HAP guidelines published by the American Thoracic Society (ATS) and by The University Health Network, Toronto, Ontario.
DESIGN: A retrospective assembly and chart review.
SETTING: A university teaching hospital.
PATIENTS: One hundred fifteen consecutive patients who had been diagnosed with pneumonia more than 48 h after admission to hospital over a 10-month period.
RESULTS: The charts of 115 patients were reviewed. Seventy-five patients (65%) were treated empirically. Forty patients (35%) were treated based on microbiological data that were available before the initiation of antibiotics. Patients who received nonempirical treatment for HAP had a significantly greater acuity of illness than the empirically treated group. Thirty-seven patients (49%) who received empirical therapy were treated according to either ATS or hospital guidelines for HAP. The use of guideline-concordant antimicrobial therapy had no measurable effect on in-hospital mortality (eight of 37 patients [21.6%] versus seven of 38 patients [18.4%], P=0.96) or median length of stay (19 days versus 21 days, P=0.30). Patients whose treatment did not follow guideline recommendations tended to receive appropriate antimicrobial coverage more often than did those patients whose treatment was guideline concordant (15 of 18 patients [83%] versus six of 11 patients [55%], P=0.49).
CONCLUSIONS: Institutional and ATS guidelines for the empirical treatment of HAP are less widely used than would be predicted by chance at The University Health Network. The clinical utility of these guidelines remains to be proven.