Guideline-Driven Antibiotic Selection Lowers Pneumonia Mortality

Home>
Specialty Care>
Emergency Medicine>
Summary and Comment

Guideline-Driven Antibiotic Selection Lowers Pneumonia Mortality

Findings of two retrospective cohort studies support the antibiotics recommended in the current national CAP guidelines.

National guidelines for inpatient care of community-acquired pneumonia (CAP) recommend administration of a fluoroquinolone alone or a ß-lactam plus a macrolide within 4 hours of hospital arrival and administration of pneumococcal and influenza vaccines. In two separate, retrospective, observational cohort studies in the San Antonio area, investigators evaluated the effects of guideline adherence with respect to the correct choice of initial antibiotics — but with a more liberal window (24–48 hours) — on process and outcome measures.

In a pharmaceutical company–sponsored study of CAP admissions to five community hospitals from 1999 through 2000, the authors measured the association between guideline-concordant antibiotic therapy (defined as administration of a recommended antibiotic within 24 hours of admission) and time to clinical stability, time to switch to oral antibiotics, length of hospital stay, and inpatient mortality. Of 631 patients, 57% received guideline-concordant antibiotic therapy. In the concordant group compared with the discordant group, time to switch to oral therapy (4.5 vs. 5.9 days) and length of hospital stay (5.0 vs. 6.2 days) were significantly shorter, and inpatient mortality was significantly lower (3% vs. 7%). No difference in time to stability was noted between groups. Only 67% of patients received antibiotics within 4 hours of arrival, and fewer than 3% were vaccinated appropriately in the inpatient setting.

In another study of CAP admissions to two academic hospitals from 1999 through 2002, the authors measured the association between guideline-concordant antibiotic therapy (here defined as administration of a recommended antibiotic within 48 hours of admission) and inpatient mortality at 48 hours. Of 787 patients, 625 (79%) received guideline-concordant therapy, and 20 patients died within 48 hours. In a multivariate, controlled analysis, mortality was lower in the concordant group (odds ratio, 0.37). Of note, 63 patients (8%) had positive blood cultures, with 34 cultures growing Streptococcus pneumoniae. Only 28% of patients received antibiotics within 4 hours of presentation.

Comment: Surprisingly, the authors chose a 24- to 48-hour window, instead of applying the guidelines’ 4-hour standard. Still, these results suggest that a fluoroquinolone alone or a ß-lactam plus a macrolide improves outcomes relative to a ß-lactam alone. However, an important potential bias exists: The most common nonconcordant regimen in both studies was ß-lactam monotherapy. Patients receiving this regimen possibly had focal infiltrates on X-ray, suggesting a "typical" pneumonia, while those receiving guideline-concordant therapy might have had small or patchy infiltrates, indicating an "atypical" pneumonia with lower mortality. If this is true, then a cephalosporin alone might be adequate for many patients. Nevertheless, an oral fluoroquinolone alone is simplest and most likely to be safe and effective. Its oral bioavailability is excellent, and it covers all bacteria that commonly cause CAP. If methicillin-resistant Staphylococcus aureus, anaerobes, or pseudomonas are suspected, broader coverage is required.

— Daniel J. Pallin, MD, MPH

Published in Journal Watch Emergency Medicine November 17, 2006