Skip to main content Back to Top


Conference Tackles Overprescribing of Antimicrobials

Kate Traynor

Changing the prescribing habits of physicians is a key part of any strategy to control the problem of antimicrobial resistance, says Neil O. Fishman, director of healthcare epidemiology and infection control at the University of Pennsylvania Medical Center (UPMC) in Philadelphia.

Fishman, speaking at the National Foundation for Infectious Diseases (NFID) 2004 Conference on Antimicrobial Resistance in Bethesda, Maryland, held June 28–30, described how UPMC relies on formulary management to influence prescribing.

Formulary control. A dozen years ago, Fishman said, UPMC instituted a restrictive antibiotic control program as part of a complete overhaul of the hospital's formulary. For each therapeutic class, UPMC selected just one drug to become part of the formulary.

"We evaluated drugs with respect to therapeutic efficacy, and if the drugs were therapeutically equivalent, then we put the least expensive agent on the formulary," Fishman said. "Increasingly," he added, "we are taking resistance issues into consideration" when making formulary decisions.

To assist with prescribing decisions, Fishman said, UPMC has "developed guidelines for antibiotic use and empiric antimicrobial therapy" that incorporate "appropriate dosing and dosage intervals based on disease state pharmacokinetic principles."

For example, he said, UPMC's guidelines recommend that cefazolin for injection be administered at a dosage of "500 mg every 8 hours in most settings and a gram every 8 hours for bacteremia," although the product's labeling recommends a dosing interval of 6, 8, or 12 hours, depending on the infection.

Fishman said UMPC has collected data to compare the antimicrobial administration guidelines with physicians' usual practices and found that the guidelines led to a higher cure rate and a lower failure rate for infections. He said there was also a nonsignificant trend toward reduced emergence of resistance when clinicians followed the guidelines.

Such data, Fishman said, are important for decision-making by physicians and hospital administrators. Also important is the associated cost savings. Fishman said UPMC's antimicrobial control program saves the hospital about $4.2 million annually in infection-related costs. Most of the savings, he said, results from a decreased length of stay in the intensive care unit (ICU) by patients whose treatment follows the antimicrobial-use guidelines.

Prescribing in the trenches. Fishman noted that decisions about drug therapy in the hospital are often made by those with the least amount of experience treating patients.

"At our hospital," he said, "it's the interns making the decisions. And as of last week, they're essentially medical students, because they just started their residencies and they're being thrust with the responsibility of making antibiotic decisions with limited knowledge" of how to best treat patients.

About half of all U.S. patients receive antimicrobial therapy during their hospital stay, Fishman said. Ideally, he added, clinicians will tailor a patient's antimicrobial therapy on the basis of microbiological data obtained from individual patients. Unfortunately, he said, "we know they don't tailor therapy."

"The clinical pharmacists that I work with actually go out every day and review charts of people that are started on broad-spectrum therapy," Fishman said. "They communicate directly with physicians to let them know about culture results and talk to them about decreasing therapy" when appropriate.

The VA experience. At the Veterans Affairs (VA) Medical Center in Pittsburgh, Pennsylvania, Victor L. Yu, chief of the infectious disease section, said physicians have voluntarily changed their antimicrobial prescribing habits, resulting in a dramatic drop in resistant microorganisms at the medical center's ICU.

"We're in the midst of an epidemic of Clostridium difficile colitis nationally, but not at the Pittsburgh VA," Yu said. "We have not seen a case of Acinetobacter or Stenotrophomonas maltophilia for six months." He said that Candida glabrata infections have also vanished, and there are fewer methicillin-resistant Staphylococcus aureus infections.

Yu credits the demise of problematic pathogens to the influence of a study that was conducted at the medical center's ICU and published in 2000.1

As the study was underway, "the doctors themselves, the ones who order excessive antibiotics, suddenly realized that maybe what they had been doing all these years was wrong," Yu told NFID conference attendees. "They curtailed their prescribing habits because of the realization that their practice not only lacked a detectable benefit but was causing measurable harm" to patients.

What physicians had been doing, Yu said, is prescribing empirical therapy for clinically suspected but unconfirmed pulmonary infections in ICU patients. In the majority of cases, Yu said, patients had pulmonary edema but no infection.

"They don't want their patients to die," Yu said about the hospital's physicians. "It's their unwillingness to risk missing a treatable infection" that results in overprescribing and subsequent antimicrobial resistance.

Persuasion, not restriction. Yu said that instead of telling physicians not to prescribe antimicrobials, the hospital devised a three-day antimicrobial regimen and compared it with physicians' usual practices, which Yu described as "blast[ing] everybody with every conceivable antibiotic" regardless of whether an infection exists.

In the study, ICU patients with a clinical pulmonary infection score (CPIS) of 6 or less, indicating a low likelihood of pneumonia, were randomized to the usual practice group or the experimental therapy—400 mg of ciprofloxacin intravenously every eight hours for three days.

At the end of three days, patients in the monotherapy group were evaluated and ciprofloxacin was discontinued if the CPIS score remained at 6 or less and microbiological test results indicated no pathogen was present. Patients in the usual therapy group continued to receive therapy at their physician's discretion.

After 81 patients had completed the study, Yu said, "we found that the critical care specialists and the surgeons were giving fewer and fewer antibiotics," Yu said. The duration of therapy in the usual therapy group, he said, "was approaching three days."

Because the study was not blinded, Yu said, physicians who treated patients in the ICU were able to see that patients randomized to receive three days of ciprofloxacin were doing better than patients receiving usual care.

Specifically, he said, antimicrobial resistance and the development of so-called superinfections were less frequent in the three-day therapy group.

From a clinical perspective, Yu said, "there was no patient who got randomized to the three-day group whose pulmonary infiltrate actually got worse." An unexpected finding, he said, was that "the patients in the three-day monotherapy group got out of the ICU significantly faster" than those who received standard therapy.

Finally, 30-day survival was similar in both groups, with a trend toward higher survival among patients in the monotherapy group. The study was halted early by an institutional review board after a VA physician declared that three-day monotherapy was his new standard therapy for ICU patients with a CPIS score of 6 or less.

"In our ICU now," Yu said, "we have found that probably over 75% of the patients don't have to get multiple antibiotics for many days." He noted that the hospital now uses levofloxacin, which costs less than ciprofloxacin, for monotherapy.

In the years since the monotherapy protocol was implemented, Yu said, no patient with a CPIS of 6 or less who received the short-course antimicrobial treatment has died from an invasive infection.

Spreading the word. Yu encouraged conference attendees to replicate the study's findings at their own institutions.

"If you replicate the study, we predict that your physicians will see what we saw," Yu said. He added that physicians who complete a rotation at the medical center "all remarked . . . how their antibiotic prescription habits have changed after coming into the VA."

"They suddenly realized that they've been overprescribing," and they voluntarily changed their prescribing habits, he said.

1. Singh N, Rogers P, Atwood CW et al. Short-course empiric antibiotic therapy for patients with pulmonary infiltrates in the intensive care unit. A proposed solution for indiscriminate antibiotic prescription. Am J Respir Crit Care Med. 2000; 162:505-11.