King of the HAI hill

Clostridium difficile infection has overtaken MRSA as the top hospital-acquired infection.

In the wake of recent reports that Clostridium difficile (C. diff) had overtaken methicillin-resistant Staphylococcus aureus (MRSA) as the top hospital-acquired infection, the time was ripe to review C. diff treatment and management at the Society of Hospital Medicine's annual meeting.

The main symptoms of C. diff are diarrhea and colitis, and they are often preceded by antibiotic exposure, noted session speaker John Bartlett, MACP, a professor of medicine at Johns Hopkins University School of Medicine in Baltimore. As such, it's often difficult to distinguish C. diff from other kinds of diarrheal illness, he said.

“If you send 100 (diarrheal) stools to a lab for C. diff testing, 80% to 85% of them will be negative. So how do you tell if it's C. diff?” Dr. Bartlett asked.

With antibiotic-associated diarrhea, the symptoms get better when the antibiotic dose is lowered. If the drug is stopped, the symptoms go away. With C. diff, stopping the drug only works about 30% of the time. Also, C. diff often carries distinct signs of colitis, like cramps, fever and leukocytosis, he said.

“Leukocytosis is a characteristic of the disease, and a barometer of the severity of the disease,” Dr. Bartlett said.

Another clinical feature? A distinctive, sickeningly sweet stool odor that some have likened to horse manure. Dr. Bartlett uses the odor as a tool for diagnosing patients with diarrhea who previously had C. diff, to help determine if they might be having a relapse. “If they report that the smell is back, then that sends a signal. And if they say the smell was there before but it's not the same this time, then it's probably not a real relapse,” he said.


Fluoroquinolone and cephalosporin are the number one and two culprits in terms of antibiotics that cause C. diff, Dr. Bartlett said. Such inducing agents clearly need to be stopped in the patient with suspected C. diff, and a different antimicrobial used to treat whatever condition necessitated an antibiotic to begin with.

The “big three” inducing agents, which should be avoided as alternative antibiotics for a patient with C. diff, are clindamycin, broad-spectrum beta-lactams and fluoroquinolones. Non-inducing agents are urinary antiseptics, sulfonamides, and any agents without antibacterial activity. If it's not possible to pick a non-inducing agent, the following carry a lower induction risk than the “big three”: narrow-spectrum beta-lactams, macrolides, tetracyclines, aminoglycosides, anti-staph drugs, metronidazole, trimethoprim/sulfamethoxazole (TMP/SMX), vancomycin and rifampin.

Some patients may not like certain of these alternative drugs, based on past experience, Dr. Bartlett said. “I give every patient this list” of the various inducing agents, and they can say what they like and don't like, he said.

Still, virtually any antibiotic can cause C. diff. Dr. Bartlett recalled the story of a 25-year-old woman who got the infection after being treated for pelvic inflammatory disease with oral metronidazole and vancomycin. “She got C. diff from the only two drugs that are used to treat it!” he said.

The next step in management is to discontinue and avoid using antiperistaltic agents for 10 to 14 days, due to concern about ileus, Dr. Bartlett said.

Once these steps are taken, it's time for a stool test. Enzyme immunoassay (EIA) is used in 95% of labs in the U.S. It is fast and cheap and 98% specific, but only 70% to 80% sensitive, “and it doesn't do any good to get two of them,” Dr Bartlett noted.

With 99% sensitivity and 98% specificity, the real-time PCR test for the toxin B gene is “the new kid on the block—and if your lab does it, it's probably your best current test,” he said. It detects the bug itself, not the toxin, so it's important that clinical symptoms are present, too, he added.

Another option is using a hybrid test. Glutamate dehydrogenase (GDH) detection, combined with the EIA, has the high specificity of the latter and the 80% to 98% sensitivity of the former. These two, plus the PCR, provide an answer in about an hour, he said.


Serious disease from C. diff is defined as a white blood cell count of more than 15,000, more than 10 bowel movements per day, and the presence of pseudomembranous colitis and severe pain. Vancomycin is the drug of choice for treatment: 125 mg by mouth four times a day for 10 to 14 days.

Moderate disease calls for 250 mg of metronidazole by mouth four times a day for 10 to 14 days, while mild disease calls for no treatment. About a third of patients get better with no treatment—although “most of us treat almost anyone with C. diff,” Dr. Bartlett said.

Resistance isn't a problem with either drug, and the relapse rate is about the same for both at 20%. Metronidazole is mainly favored for moderate disease because it's lower in cost than oral vancomycin.

“In the hospital we give vancomycin in the IV form for the patient to drink, and it's a fraction of the cost. Every once in a while you will find an outpatient pharmacy that will formulate it that way, and therefore the price for a ten-day supply goes from $900 to $90. Metronidazole is a little cheaper than that,” Dr. Bartlett said.

For serious cases and as a last resort, stool transplants are almost always successful, Dr. Bartlett noted. The ideal person to do it is the radiology technician who does barium enemas: He or she knows how to access the right colon.”I don't know how many of you have done (a stool transplant), but pathophysiologically it's a dream,” Dr. Bartlett said. “Aesthetically, it sucks.”


Relapse has long been a challenge with C. diff. Happily, the same treatment Dr. Bartlett prescribed to his patients back in 1978 still works well today—a tapering dose of vancomycin, followed by a “pulse” dose, as follows:

  • 125 mg of vancomycin four times a day for 10 to 14 days, then
  • 125 mg of vancomycin twice a day for seven days, then
  • 125 mg of vancomycin once a day for seven days, then
  • 125 mg of vancomycin every other day for four to six weeks.

Why does pulsing work on relapses? The theory—which hasn't been tested but is nonetheless in treatment guidelines—goes like this: The tiny pulse dose of vancomycin is just enough to keep the bug a spore, but not enough to alter the competing gut flora. While the C. diff is quiet, the normal flora has a chance to return over the course of four to six weeks.