Managing serious infections in the hospital: a new model

Abstract

The purpose of this symposium was to consider the new approaches to the management of serious infections of hospitalised patients. The symposium focused on two areas, febrile neutropenic patients and ventilator-associated pneumonia. However, the issues of antibiotic resistance are relevant for many other areas in the hospital. One of the major issues facing the treating physician is the lack of new drug development. This is especially true for the Gram-negative bacilli. For various reasons, the number of drugs being developed has been sharply reduced over the past few years. This reduction has led to a more urgent need for clinicians to learn to do more with what they have. The various participants in this symposium have therefore provided help for the clinician to do more with the drugs we have available. Dr Peterson pointed out that increased drug usage, with resulting increased exposure, can be expected to lead to resistance. While this has been far more rapid with some antibiotics, such as cephalosporins and fluroquinolones, one could expect it with all antibiotics. In one study, the use of imipenem as a replacement for a cephalosporin was associated with high levels of resistance for Pseudomonas aeruginosa within a year [1]. In some ways, this can be compared to squeezing a balloon; restricting one antibiotic leads to another antibiotic losing its effectiveness [2]. However, prolonged use of piperacillin–tazobactam for empirical treatment has not been associated with the same rapid rate of emerging resistance. Aminoglycosides are another example of antibiotics which lead to resistance only after prolonged exposure. However, switching from gentamicin to tobramycin led to recovery of the effectiveness of gentamicin [3]. The use of the piperacillin–tazobactam or an aminoglycoside for empirical therapy may allow one to resuscitate some of the other antibiotics. Dr Glasmacher has demonstrated how metaanalysis allows for pooling information from previous studies. This enables the clinician to answer questions which were not apparent from individual studies. For example, meta-analysis of studies of the febrile neutropenic patient demonstrated that monotherapy was safer than combination therapy with aminoglycoside. In addition, monotherapy was at least as effective. Although recommendations from such studies are useful, they may not apply at a particular hospital. If there is a large problem with multidrug resistance, aminoglycosides may still be needed for treatment of neutropenic fever. However, the clinician has the evidence to support the use of monotherapy for the majority of patients with neutropenic fever [4]. Dr Bow also used meta-analysis to compare new study results with those of prior trials. He showed the benefits of combined analysis; these included demonstration that the failure rates of some of the new therapies may be as high as those found in prior studies. However, he pointed out that the overall mortality was lower with current therapy. This demonstrates that before one changes empirical therapy, one has to be sure which outcome is most important. He proposed mortality as the standard to evaluate therapy. The importance of knowledge of local flora was discussed by Dr Bow. Fluoroquinolones have been found to be useful in some studies in preventing neutropenic fever. Analysis of studies to date demonstrates that this strategy is effective as long as the rate of Gram-negative bacillary resistance is less than 10%. If the rate of resistance is higher, then there is no advantage with fluoroquinolone therapy. The outcome of ventilator-associated pneumonia (VAP) is also affected by the rise of multidrug-resistant bacteria. The three most common bacteria causing VAP are Staphylococcus aureus, P. aeruginosa, and Acinetobacter baumannii [5]. All of these are associated with increasing rates of resistance. Antibiotic strategies must include ways in which to reduce the rate of resistance during therapy. In addition, the effects of antibiotic therapy on other bacteria need to be considered.