Antibiotic restriction policies: are they for you?

May 2004

Two papers published recently in Clinical Infectious Diseases are particularly worth reading, especially if your hospital is considering placing restrictions on the use of certain antimicrobial agents. The first, by Sunenshine et al, is a survey describing the role of infectious disease consultants carried out within the Emerging Infections Network (EIN) of the IDSA (2004;38(7):934-938). The second is an editorial by John McGowan of Emory University, who has long been an observer and commentator on controlling antimicrobial resistance in hospitals (2004;38(7): 939-942).

The survey consisted of a two-page introduction and a one-page questionnaire sent to all 690 EIN members. Nonresponders received one reminder two weeks later. The overall response rate was 73%, which should be considered surprisingly good and undoubtedly a testimony to the brevity of the survey. The three general areas of inquiry were the relationship between antimicrobial use and resistance development, the role of ID consultants in hospital approval processes if restrictions were in place, and acceptance – or lack of acceptance – of that role by other staff physicians. The survey was carried out in March 1999, and one wonders how the findings may have evolved over the last five years.

Overall, 50% of respondents indicated that their hospital required approval of certain antibiotics by an ID consultant prior to release. Perhaps not surprisingly, the proportion of hospitals that had restrictions varied considerably according to the type of hospital: restrictions were in place in 70% of university teaching hospitals, about 50% of nonuniversity teaching hospitals and only 17% of nonteaching hospitals. Concerns about referral patterns and effect on consultations, ie, the “policeman effect,” were just the reverse: those concerns were most marked in the nonteaching hospitals and less in teaching hospitals. There was also some geographic variation in the use of restrictions, ranging from a high of 68% in the New England and mid-Atlantic regions to a low of 23% in the West-North Central region.

The specific antimicrobial agents restricted varied somewhat by hospital, but the most commonly restricted agents included lipid formulations of amphotericin B, carbapenems, fluoroquinolones, piperacillin-tazobactam (Zosyn, Wyeth) and vancomycin. It seems that both cost and control of resistance development entered into the restriction decisions.

Most surprising was the finding that only 18% of the 250 respondents in hospitals with restrictions in place reported that ID consultants were directly remunerated for their participation in the process! A few, however, reported “indirect” benefits such as funding for fellowship positions, increased consultation requests and the like.

In the accompanying editorial, McGowan took a broad view of the problem and the possible solutions. He saw the role of the ID physician as critically important in three areas:

Education and example-setting in facilitating compliance — especially at the medical staff level — with infection control procedures; failures in infection control regularly spread multiresistant organisms in health care settings.

• ID physicians, particularly through their peer organization, the IDSA, must continue to bring the steady erosion of pharmaceutical manufacturers involved in developing new antibacterial drugs to the attention of government and the public.
• Most important of all, judicious use of antimicrobial drugs remains the cornerstone for limiting development and spread of multiresistant organisms in hospitals.

There is by this time abundant evidence that infection with multiresistant organisms is more expensive to treat and that the incremental cost of infection by such organisms is, for the most part, not reimbursed by insurers. Rather, it is borne by the hospital itself, resulting eventually in increased charges for all patients. Many reports have been published that estimate the savings to be realized by antimicrobial “control” programs, and there are finally beginning to appear actual results of such programs. In a hospital such as my own, a 350-bed tertiary care center, we estimated that a control program involving the antibiotics mentioned in the Sunenshine report could bring about an annual savings of $250,000! This would save far more than the cost of the program itself.

Physician acceptance remains the largest challenge, as demonstrated in the Sunenshine report. Hospital administration “buy-in” and medical staff “buy-in,” especially at the leadership level, are critical, and any control program that lacks this level of support is likely to fail. It is terribly difficult for a physician, acting in the perceived best interest of a patient, to appreciate the link between what he or she prescribes at that moment in time and the emergence of antimicrobial resistance in the future. If antibiotic restriction decisions are seen as wholly arbitrary, there will be rebellion by the medical staff. Thus, there is a huge challenge to educate medical staff so that they can appreciate the multidisciplinary nature of restriction recommendations and that they are based, insofar as possible, on sound data.

Another VRSA

Also on the antimicrobial resistance front, at the recent meeting of the Society for Healthcare Epidemiology of America (SHEA), the CDC announced the third isolate of vancomycin-resistant Staphylococcus aureus (VRSA) in the United States (also published in MMWR. 2004;53[15]:322-323).

The organism was recovered from the urine of a resident of a long-term care facility. One hopes that more clinical details of this patient will emerge, so that the virulence of that strain can be better evaluated. No transmission of the organism is known to have occurred as yet.

The important point made in the MMWR report is that the presence of vancomycin resistance was not detected by a commonly used overnight automated antimicrobial susceptibility test, Microscan (Dade Behring). That test reported a minimum inhibitory concentration (MIC) of 4 µg/mL, but the organism was later found by E-test (AB Biodisk North America) to have an MIC of >256 µg/mL. Subsequent testing at the CDC using the National Committee for Clinical Laboratory Standards microdilution reference method determined the MIC to be 64 µg/mL. The important take-home message for clinical laboratories is to use a vancomycin-agar screening plate if automated overnight susceptibility tests are used.

This report is a none-too-gentle reminder that VRSA is still with us and will undoubtedly reappear from time to time. Judging by our experience with methicillin-resistant S. aureus, which appeared first in the United States in 1968 but did not really turn into a major national epidemic until the late 1980s and 1990s, we can expect VRSA to pop up sporadically for perhaps five to 15 years before literally exploding among us. We would do well to heed the lessons of history and prepare for those days that surely lie ahead.

For more information:

• McDonald C. Third case of vancomycin-resistant Staphylococcus aureus. Presented at the 14th Annual Scientific Meeting of the Society for Healthcare Epidemiology of America. April 17-20, 2004. Philadelphia.