© 2004 American Thoracic Society
Antimicrobial Treatment of Ventilator-associated PneumoniaTo the Editor:I read the article by Ost and colleagues (1) and the accompanying editorial by Drs. Amin and Hébert (2) with great interest. Ost and colleagues concluded that patients with ventilator-associated pneumonia should be treated with three antibiotics with mini–bronchoalveolar lavage from the perspective of minimizing cost, minimizing antibiotic use, and maximizing survival. Emergence of resistant strains is an increasing problem of intensive care units and probably is the result of using broad-spectrum antibiotics. Immediate cost analyses generally do not include this parameter. The design of the decision analysis does not allow us to assess whether treatment leads to antimicrobial resistance. Growing data suggest that one of the most important factors in the development of resistance is antibiotic use. Three antibiotics given empirically might represent excessive use of antibiotics. Areas within the hospital with the highest resistance rates have the highest antimicrobial use, and increased duration of patient exposure to antimicrobials increases the likelihood of colonization with resistant organisms (3). The names or groups of these three antibiotics should be mentioned clearly in the article. The conclusion of the article may mislead physicians to use unnecessarily antibiotic combinations that most probably result in treatment failure, emergence of antibiotic resistance, and antagonism.
Ondokuz Mayis University Medical School Samsun, Turkey FOOTNOTES Conflict of Interest Statement: H.L. has no declared conflict of interest. Dr. Hébert was given an opportunity to respond to this letter but declined to do so. REFERENCES
Conflict of Interest Statement: D.O. and A.F. have no declared conflict of interest.
From the Authors: Dr. Leblebicioglu addresses two important issues: the long-term impact of antibiotics on the emergence of resistant strains and optimal antibiotic selection. Dr. Leblebicioglu also notes "the design of the decision analysis does not allow us to assess whether or not treatment leads to antimicrobial resistance." Although this is correct, it was not the question being asked in this study (1). Indeed, there is already a large body of evidence that suggests antibiotic use often does lead to resistance over the long-term. The real issue is how to balance this with other competing goals. The competing objectives we need to consider are the need for rapid, effective treatment to maximize survival, the need to minimize financial cost, and the need to avoid unnecessary antibiotic use. To systematically investigate the tradeoffs involved in the clinical decision process was the purpose of this study. Our analysis does not directly address the impact of using antibiotics on long-term resistance rates (1). As noted in the DISCUSSION, this was not feasible because there is insufficient evidence to precisely model the relationship between antibiotics used today and the future probability of antibiotic resistance. However, as noted in the RESULTS, the model does measure antibiotic days per survivor, therefore cost-effectiveness techniques can be used to gain insights into the problem. Specifically, it allows us to more systematically ask the following question: How much benefit are we gaining (in terms of survival and financial cost) in our use of antibiotics? The second question raised is: what specific antibiotic(s) should be used? However, as noted in the DISCUSSION, there is significant variation in local incidence and resistance patterns, both among hospitals and among intensive care units (ICUs) (2, 3). What is the optimum antibiotic(s)? The answer will depend on local factors, specifically the local antibiogram and the local incidence patterns. As an example, the "correct" answer for hospitals with a high incidence of Methicillin-resistant Staphylococcus aureus (MRSA) will differ significantly from those with a low incidence of MRSA (2). It is not the number of antibiotics alone that is important, but rather the probability of adequate coverage that results for a given number of antibiotics that is of importance. The key is quantification of the incremental benefit of using more than one antibiotic. Thus, if one antibiotic allows 98% coverage, the incremental benefit of a second antibiotic cannot be more than 2%. However, in late-onset ventilator-associated pneumonia (VAP), single-agent therapy is often ineffective (4). Because there is currently no mathematic model available that would be broadly applicable across ICUs to make this calculation, we used data from previous studies to determine the relationship between number of antibiotics and probability of adequate coverage (see online supplement to our article) (1). If the probability of adequate coverage falls within the assumptions of the model, then broader initial coverage with rapid narrowing will be more cost-effective. Note that antibiotic regimens that are too narrow will be ineffective and costly, because they often will have to be changed, these changes will come too late to alter outcome significantly, and they will still promote resistance. In essence, from a cost-effectiveness perspective, the most expensive antibiotic is the one that does not work. This is not to say that antibiotics should be used indiscriminately. As our analysis suggests, overuse can be minimized by adjusting antibiotics on the basis of diagnostic tests and limiting the duration of use. Studies published subsequent to our own suggest that perhaps limiting the duration of therapy to 8 days may be as effective as 14 days (5). This alone would be of great benefit in terms of limiting antibiotic-days. Limiting antibiotics is part of a more complex multidimensional objective, but it is not the sole objective. This needs to be considered when developing a strategy for VAP. As with most issues of strategy, it is striking a balance that is critical. In this case it is between overuse (leading to unnecessary resistance in the future) and underuse (leading to unnecessary mortality and financial cost in the present).
North Shore University Hospital Manhasset, New York REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
