Surgeons need help with antibiotics

Most surgeons misuse antibiotics most of the time. That is the blunt warning from Robert Sawyer, M.D., codirector of the surgical trauma intensive care unit at the University of Virginia Health System. The problem is not ignorance of appropriate use, he told the American College of Surgeons Clinical Congress meeting in San Francisco recently.

The problem is that clinical practice has lagged behind clinical knowledge. "Knowledge alone does not alter behavior," Sawyer said. "We've known for 30 years what should be done with antibiotics in surgery, and it's not happening. We need to be checked by pharmacy, anesthesiology, and nursing to ensure good antibiotic practice."

Almost every area of antibiotic use in surgery needs improvement, agreed Addison May, M.D., associate professor of surgery and anesthesiology and director of the Trauma/Surgical Critical Care Residency at Vanderbilt University Medical Center. Nearly 40% of surgical patients are given presurgical antibiotic prophylaxis either too early or too late. A significant percentage of patients are given a broad-spectrum agent when they should receive a narrow-spectrum agent. Other patients are given antibiotics in the absence of evidence of infection. Even patients for whom an antibiotic is appropriate are given too long a course.

Multiple studies agree that pre-op antibiotic prophylaxis is most effective when administered one hour before the first incision. Yet a recent study across 44 teaching hospitals found that 14% of patients received no antibiotic prophylaxis at all and 37% got their presurgical dose at the wrong time.

Why the concern? The surgical infection rate among patients who received prophylaxis within an hour of first incision was 1%, May noted. The infection rate jumped to 4% when prophylaxis was given more than two hours before surgery began.

"The consensus is clear that prophylactic antibiotics are useful in reducing surgical infections," said Thomas Gleason, M.D., director of aortic surgery at Northwestern Mem- orial Hospital in Chicago. Single-dose prophylaxis is just as effective as multiple doses, at least in cardiac procedures. There are few, if any, circumstances under which prophylaxis should continue following surgery, he added.

One reason for limiting the use of prophylactic antibiotics is the relationship with antibiotic exposure and later nosocomial infection such as ventilator-associated pneumonia. Exposure to any antibiotic within the prior seven days boosts the odds ratio of acquiring nosocomial pneumonia to 13:46, May noted. Antibiotic use also exerts strong selective pressure in favor of drug resistance, he added. Widespread use of vancomycin, third-generation cephalosporins, and fluoroquinolones is strongly associated with the rise of drug-resistant pathogens in hospitals.

The desire to stop potentially resistant infections quickly is an important reason patients who present with a possible infection are often treated empirically. Empiric therapy is appropriate for many patients, but appropriate drug choice is critical. Too many clinicians order broad-spectrum products in the mistaken belief that it is better to attack more bugs than fewer, May cautioned. The reality is that narrow-spectrum agents are more effective and less likely to induce resistance.

A two-year study of patients with an implanted central port found that the use of broad-spectrum antibiotics was associated with a 4% infection rate. Switching to narrow-spectrum agents cut the rate to 1.8%. Studies in the mid-1990s found that use of broad-spectrum agents increased the incidence of methicillin-resistant Staphylococcus aureus (MRSA) by three to four times.

Other studies have found patients on antibiotics even though they did not have an infection, cautioned Soumitra Eachempati, M.D., associate professor of surgery and public health at Cornell University's Weill Medical College. Surgeons saw a fever, assumed infection, and wrote an antibiotic order. "You've got to look for the source of fever," he said. "They are not all susceptible to antibiotic treatment."

Length of treatment is another problem. Fully 24% of adverse drug events are related to antibiotic use, Sawyer said. Long-term use of antibiotics is associated with higher mortality. Prior antibiotic use is also one of the most important risk factors for fungal and candidemia infection.

"If a little antibiotic is good, more is not better," Sawyer said. "Patient exposure to antibiotics should be as limited as possible."