Antibiotic Resistance—Squeezing the Balloon?

The need for responsible antibiotic use stewardship to quell microbial resistance should have no disputants. Indeed, there have been so many clarion calls for action¹ to halt the increasing resistance to antimicrobials that further emphasis seems redundant. The list of august organizations and agencies that have highlighted the problems of antibiotic resistance in recent years includes the National Academy of Sciences, Institute of Medicine, the Infectious Diseases Society of America, the Centers for Disease Control and Prevention; the American Society for Microbiology, and the World Health Organization. The National Foundation for Infectious Diseases has ranked antimicrobial resistance and emerging infections as the first among the top 10 problems in infectious diseases on which it will concentrate its efforts.²

The unprecedented level of concern about increasing drug resistance, both in the community and in hospitals and involving a wide variety of human pathogens, reflects the growing frustration with the intractability of this multifaceted problem. Most of the consensus recommendations for preventing antibiotic resistance include a call for practice guidelines and other institutional policies to control the use of antibiotics. On the other hand, physicians often seem to lack commitment to these objectives especially when the needs of individual patients may seem in jeopardy. There has even been concern that managed care may be inadvertantly contributing to the problem by encouraging overprescribing because of heavy clinician workloads and all-too-brief patient visits and, perhaps, even by limiting relevant (and costly) microbiologic culture testing that could otherwise help curtail the use of broad-spectrum antimicrobials. Even experts cannot agree, it must be admitted, on successful strategies. After a formal "Oxford" style debate, members of the Hospital Infection Society in the United Kingdom voted nearly equally for and against the motion: "This house believes that hospital antibiotic policies have no effect."³

Why is this issue contentious? It seems clear that use of antimicrobials encourages the development of resistance in bacterial strains. However, the genetic mechanisms used by bacteria to acquire antibiotic resistance not only promote dissemination of resistance but also favor stability of resistance genes even in the absence of ongoing exposure to the antibiotic.⁴ Population geneticists, therefore, have concluded that technology is losing the arms race with evolution.⁵ However, excessive prescribing must still be avoided to prevent further dissemination of such genes. Therefore, the report by Rahal and colleagues⁶ in this issue of THE JOURNAL that restriction of cephalosporin use was followed by a reduction in the frequency of β-lactam resistance amongKlebsiella species should be welcomed as good news. The unintended consequence of the hospital's restriction program was an increase in use of imipenem and a concomitant increase in imipenem-resistant Pseudomonas aeruginosa. As medical cost control has been compared to squeezing a balloon—constraining one end causes the other end to bulge—addressing the problem of antibiotic resistance by limiting the use of one class of compounds may be counteracted by corresponding changes in prescribing and drug resistance that are even more ominous.

The perverse and unintended consequences of "component management" have been recognized in other areas of formulary management in which modest cost savings have been offset by unexpected increases in adverse drug reactions and high costs for other hospital services.⁷ In this context, antibiotic resistance may be viewed as another type of adverse drug event. To date, antibiotic restriction programs have had disappointing results, if only because of the costs for their implementation and the difficulty in sustaining such efforts over time. Kunin⁸ has underscored the bleak outlook for control of antibiotic resistance by such means when he noted, "There is only a thin red line of infectious disease practitioners who have dedicated themselves to rational therapy and the control of hospital infections."

Strategies that aim at optimizing antibiotic management may be most able to avoid squeezing the balloon. However, education, including "academic detailing,"⁹ and top-down policymaking alone are insufficient. Antibiotic cycling, or rotation among various antibiotics of the same class, is problematic and fails to address the issue of antibiotic misuse.¹⁰ Despite reports in the literature of short-terms effects (such as in the report by Rahal et al), there is no evidence that the excessive and inappropriate use of antibiotics will be controlled by education, imposing guidelines, restrictive formularies, or other administrative measures. The reason for this may be that antibiotic misuse more often results from inadequate information than from inappropriate behavior. Antibiotics often are chosen in a setting of incomplete knowledge of causative pathogens, lack of appreciation for pharmacodynamics, and absence of expert opinion. To address this problem, information relevant to an individual patient must be provided at the bedside when a clinician is deciding to use an antibiotic.

Despite some claims of success, no single institution has succeeded in the broad effort of preventing antibiotic resistance. Outcome measures from interventions to control antibiotic use have been limited to reduced use of certain targeted drugs, changes in antibiotic susceptibility patterns of selected microorganisms, or costs of antibiotics.^{11 - 12} Many reports do not focus on the outcomes these measures have on patients, although a recent study reported decreased rates of some nosocomial infections.¹³ Scientifically sound programs that can be shown to improve the quality of patient care, not merely save costs or measure the performance of physicians vis-à-vis use of antimicrobials, are sorely needed.

Quality improvement has become synonymous with the drive to eliminate variation. However, there is a danger that standardization could lead to antibiotic prescribing patterns that could be described as homogeneous. Antibiotic restriction programs pose the danger of fostering monosynaptic decision making, ie, reflexive prescribing of favored compounds, that could heighten the potential for misuse of unrestricted drugs. There are now many examples in which the extensive use of single classes of antimicrobials have led to marked increases in resistance, such as the occurrence of ceftazidime-resistantKlebsiella pneumoniae on hospital wards where ceftazidime was administered most frequently.¹⁴ Efforts to encourage heterogeneity and individualization of antibiotic use might offer the best hope for stabilizing the selection of resistance.

The appropriate use of consultation with infectious disease specialists and medical microbiologists is an obvious technique to encourage heterogeneity through individualization of drug selection. Any method that brings more information to the bedside that will affect clinical decision making will encourage heterogeneity. Advanced computer systems can assist physicians when selecting antibiotics by providing up-to-date antimicrobial susceptibility patterns for nosocomial pathogens recently isolated from the local hospital, by displaying the costs of formulary antimicrobials, by recommending dosages and durations of therapy, by calling attention to drug incompatibilities, and even by creating guidelines for antibiotic use that are locally derived and acceptable to physicians.¹⁵ Computer-assisted decision support, based on the principles of epidemiology and using a flexible and dynamic medical informatics system, can be used to inform and improve antibiotic decisions rather than to enforce them. These patient-centered strategies may reduce the total amount of antibiotic use, promote unintentional random use of specific antimicrobials, and, thereby, stabilize antibiotic resistance. The key principle is that the individual physician retains control of antibiotic use on the basis of relevant information available at the bedside, using the computer system as a tool.

In a recent World Health Organization Symposium, Acar¹⁶ called for reexamining the relationship between resistance and antibiotic consumption, noting that "positive changes in prescribing behavior occur when information on resistant organisms is released to practitioners and the public. The impact of such information has not been investigated and measured." Within this scenario, making useful information available at the point of care and meeting the needs of physicians to use data more efficiently in their daily work should become the focal point for optimizing antibiotic use and, perhaps, reducing the selection pressure in the antibiotic balloon.