© 2005 American Thoracic Society doi: 10.1164/rccm.2506009
Does Guideline Adherence for Empiric Antibiotic Therapy Reduce Mortality in Community-acquired Pneumonia?
a University of Lausanne, Lausanne, Switzerland The single most important medical advance in the treatment of patients with community-acquired pneumonia (CAP) in the past century was the development of effective antibiotic agents in the 1940s, reducing mortality from 30 to 35% in the preantibiotic era to a current average of 5% among hospitalized and ambulatory adults with this illness (1, 2). More recently, two antibiotic-related processes of care for hospitalized patients with CAP, selection of appropriate empiric antibiotic therapy and rapid administration of such therapy, were shown to reduce short-term mortality (3–5). The recognition that large variability in clinical practices exists and that certain processes of care are associated with improved medical outcomes has led to the development of clinical practice guidelines for the management of CAP endorsed by professional organizations, such as the American Thoracic Society and the Infectious Diseases Society of America (6, 7). Despite a growing evidence base, whether appropriate empirical antibiotic therapy improves medical outcomes for CAP remains open to debate. Although several prospective and retrospective cohort studies showed a positive association between guideline adherence for initial antibiotic selection and short-term mortality (5, 8–11), others did not (12, 13). A recent meta-analysis of 24 clinical trials failed to demonstrate any survival benefit to inpatients with CAP receiving additional antibiotic coverage for atypical pathogens (14), a practice recommended by specialty society guidelines and supported by empiric observational data (5–7). It is the authors' observation that the vast majority of clinical trials for antibiotic therapy of lower respiratory infections are designed to establish equivalency of newer proprietary agents rather than establish superiority of single or combined classes of agents. As a result, the majority of such trials are small in size and do not contribute to the most relevant therapeutic questions for patients with CAP, such as will antibiotic therapy decrease mortality or hasten time to return to baseline level of functioning. In this issue, Menendez and coworkers (pp. 757–762) conducted a prospective study of 1,228 inpatients with CAP to determine: (1) which factors were associated with better adherence to Spanish guidelines for initial antibiotic selection, and (2) whether guideline adherence was associated with treatment failure or mortality (15). Overall guideline adherence was excellent (80%), but with substantial variability across study sites (range, 47–97%). Factors that were independently associated with guideline adherence were hospital site, specialty of the physician in charge (higher with a pulmonary resident or attending), and a higher severity of illness at presentation based on the Pneumonia Severity Index. Treatment in an intensive care unit was associated with lower guideline adherence. After adjusting for disease severity and physician characteristics, treatment with guideline-adherent antibiotics led to a lower rate of treatment failure (odds ratio [OR], 0.65; 95% confidence interval [CI], 0.5–0.9), and to a sizeable reduction in 30-day mortality (OR, 0.55; 95% CI, 0.3–0.9). The study by Menendez and colleagues is one of the first to assess the factors associated with adherence to guideline recommendations for empiric antibiotic therapy for CAP. Although these authors found that severity of illness and provider specialty were associated with higher levels of adherence, the clinical implications of these findings are uncertain. Severity of illness at admission is a nonmodifiable patient characteristic, and the vast majority of patients with CAP will continue to be treated by primary care and emergency department physicians and not by pulmonary or infectious disease specialists. It would have been interesting to know if there were certain provider or hospital characteristics associated with guideline adherence, particularly factors representing modifiable barriers or facilitators to appropriate prescribing practices. For example, (1) what was the role of provider knowledge and attitudes toward evidence-based guidelines for CAP? and (2) did the sites have pneumonia opinion leaders, standardized order forms, or computerized reminders to reinforce the guideline recommendations? A recent cluster-randomized trial demonstrated that an intensive guideline implementation strategy (i.e., using real-time reminders, audit and feedback, and continuous quality improvement activities) was most effective at improving adherence to guideline recommendations for empiric antibiotic therapy (16). Clearly, more research focusing on the patient, provider, and system-level barriers and facilitators of guideline adherence is an important step toward improved quality of care for CAP. The finding of Menendez and coworkers that guideline-compliant antibiotic therapy was strongly associated with improved survival suggests that the medical field is in a position to further reduce the mortality of CAP. However, the magnitude of the process–outcome association observed in this study must be interpreted with caution for several reasons. First, nonrandomized study designs, in contrast to clinical trials, are subject to confounding by unmeasured variables. Although Menendez and colleagues adjusted for physician characteristics (specialty and training status) and illness severity using the Pneumonia Severity Index, they did not adjust for other processes of care (e.g., timely administration of antibiotics, performance of blood cultures, treatment duration, and the treatment adherence after discharge), local microbial resistance patterns, or hospital characteristics (e.g., teaching status and patient volume) that may influence short-term mortality. Second, patients who died during the first 48 hours after admission were excluded from the analyses to avoid cases unlikely to be affected by the selection of antibiotic therapy. This is problematic because the exclusion of patients who may have died despite full guideline adherence may have biased the study results in favor of guideline adherence. Third, the reduction in mortality after guideline adherence was only observed in the less severely ill patients admitted to medical wards. It is difficult to understand why the beneficial effects of guideline adherence were demonstrable in patients of lower disease severity compared with the more severely ill patients who were managed in an intensive care unit. Finally, although patients treated by the same physician and within the same hospital tend to be treated similarly, the authors did not adjust for clustering at the physician or site level in their analyses, which may have decreased the 95% confidence intervals and increased the significance level of their results. We find the results of Menendez and coworkers encouraging for all who believe in the meritorious effects of practice guidelines and their association with improved patient outcomes. However, given the observational nature of studies that have attempted to quantify the magnitude of this effect, we remain skeptical that better adherence to antibiotic recommendations reduces mortality by almost 50%. In the future, methodologically sound clinical trials are required that clearly establish the association between recommended antibiotic treatments and clinically relevant medical outcomes. Such trials should establish (1) the efficacy of one or more classes of antibiotic agents by testing superiority rather than equivalency hypotheses, and (2) the effectiveness of the best methods to translate practice guideline treatment recommendations into clinical practice. FOOTNOTES Conflict of Interest Statement: D.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.J.F. received $3,500 from Aventis Pharmaceuticals in 2004 and $1,000 from Zynx Health Inc. in 2003, 2004, and 2005 for serving on an Advisory Board. He received $2,000 in 2002 from STA Healthcare Communications, Inc. for providing a lecture and received royalties in the amount of $120 per year from UpToDate, Inc. for written materials. He currently has a grant for $43,000 from Pfizer, Inc. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
