Published in the August 2013 issue of Today’s Hospitalist
WHEN IT COMES TO serious infectious diseases in the hospital “think asymptomatic bacteriuria, even with pyuria “the best advice may be, “Don’t just do something, stand there.” But for other infections like Staphylococcus aureus, the best default course may be to treat even suspected cases aggressively.
So how do hospitalists know when aggressive measures are called for? In several presentations at this spring’s Society of Hospital Medicine annual conference, infectious disease subspecialists discussed when to wait and when to run, what to start and what to stop, and why and how to definitely call for backup, both from surgeons and infectious disease experts.
Here’s a look at the mistakes those physicians regularly see their hospitalist colleagues make, and answers to the most common questions they hear.
Whatever the cause of S. aureus bacteremia, the infection is “bad news,” according to John Flaherty, MD, a professor of medicine at Northwestern University’s Feinberg School of Medicine and associate chief of its division of infectious diseases. His advice? “When in doubt, treat patients longer.”
That approach holds true whether S. aureus is methicillin-resistant or methicillin-susceptible. Because only one-half to two-thirds of patients with either MRSA or MSSA in their blood will still be alive a year after their infection, “S. aureus bacteremia is a significant event in someone’s life,” explained Thomas G. Fraser, MD, vice chair of the infectious disease department at Cleveland Clinic.
The preferred antibiotics are oxacillin or nafcillin, but cefazolin is also an option. Vancomycin or daptomycin should be used if the patient is infected with MRSA or has a bona fide penicillin allergy.
Dr. Flaherty said that persistent bacteremia “lasting more than 72 hours “may need between four and six weeks of therapy; endocarditis, six weeks; and four to eight weeks for deep tissue infections like osteomyelitis, mediastinitis or deep abscesses, depending on the patient’s response to treatment. And “never assume Staph aureus is a contaminant,” he said. If a blood culture comes back positive, treat.
When thinking about the perils of S. aureus, Carlos M. Isada, MD, Cleveland Clinic’s infectious diseases vice chair, said that while S. aureus endocarditis can occur on perfectly normal valves, people with mechanical valves or with other devices are at highest risk. He added that transesophageal echocardiograms (TEEs) are much better at diagnosing endocarditis than transthoracic echocardiograms (TTEs).
But because TEEs can be risky, costly and delay care, Dr. Isada noted that many hospitalists wonder if the test is really necessary for all patients with a positive blood culture. While TEE is recommended in the MRSA treatment guideline from the Infectious Diseases Society of America, Dr. Isada said there is ongoing research into which patients truly need the test.
One study concluded that you can reduce the number of TEEs by one-third if you limit TEEs to patients at higher risk for endocarditis, such as those with positive blood cultures lasting more than four days, patients with a permanent intracardiac device, or patients on hemodialysis or who have a spinal infection or nonvertebral osteomyelitis. “Maybe these criteria will be part of the thought process for the next iteration” of guidelines, Dr. Isada said.
Catheters and “space junk”
S. aureus is the second most common cause of nosocomial bacteremia, and the most deadly. Its incidence is rising sharply because of the growing number of immunocompromised patients, the frequency of invasive procedures and the growth of resistant strains.
If patients have a hospital-acquired S. aureus infection, there is more than a 60% chance that it will be MRSA. About 50% of health care-associated S. aureus will be MRSA, compared to 14% of community-acquired cases.
S. aureus endocarditis is bad and over time, survival is worse,” Dr. Fraser said. “Endocarditis is rapidly becoming a Staph aureus disease. When we think about heart valve infection and bloodstream infection, we have to be thinking about health care-associated infections. These are infections that usually start with catheters, and the poster child for this is the person on dialysis.”
When a “bloodstream infection is related to a catheter, Dr. Fraser said the big question is whether it’s possible to leave the catheter in place. “The takeaway from all the guidelines is if patients are sick, pull the line. If they have a bad bug, pull the line. If you try to save the line, follow them very closely,” he said. “For the bad boys, get it out.”
More than one-third of people with prosthetic joints who get S. aureus in the bloodstream will end up with a seeded joint, said Dr. Fraser. The same is true with pacemakers, defibrillators, prosthetic valves or anything else implanted, even old lines “”space junk” “left behind from previous procedures.
Moreover, he added, patients can become sick from these seeded joints long after you think the initial infection is cured. The fact that people are living longer with chronic medical illness “dovetails with more device-related infections. I think we are going to see more of this.”
If an infected line can’t come out, antibiotic lock therapy can be an option with systemic antibiotics. But, Dr. Fraser cautioned, while literature shows that it works, “I have had trouble operationalizing it, particularly sending someone home on the therapy.”
As for when to replace a pulled line, that “depends on the clinical need,” he said. “In the ICU, you may need to take it out and put another in. But I would like the blood cultures to be clear for 72 hours” or even longer for some infections.
The more experts learn about asymptomatic bacteriuria, the less they recommend treating it with antibiotics, in part because the condition is so common. Incidence rates range between 1% and 5% of healthy young women to between 25% and 50% of elderly women in long-term care facilities, and 100% of people with long-term indwelling catheters.
A study of young women found that treating asymptomatic bacteriuria actually resulted in more symptomatic urinary tract infections (UTIs) than when asymptomatic women weren’t treated.
“The theory is that asymptomatic bacteriuria is protective,” Dr. Flaherty explained. “It is essentially colonization with a relatively avirulent strain of E. coli, which is blocking the more virulent strain. You kill that off, and you will get the more virulent form. If bacteria aren’t causing any problems, leave them alone.”
A common mistake, Dr. Flaherty added, is assuming that pyuria along with bacteriuria is an indication for antibiotic therapy. “Pyuria is not diagnostic of catheter-associated UTI. You need symptoms too.” On the other hand, “the absence of pyuria in a symptomatic patient suggests a diagnosis other than a UTI.”
The current recommendation, he noted, is to refrain from screening or treating asymptomatic bacteriuria except in pregnant women or patients about to undergo urologic procedures. This includes not regularly screening diabetic women, elderly and institutionalized people, people with spinal cord injuries, or catheterized patients while the catheter remains in place.
“There is no indication for this testing,” Dr. Flaherty said, “but it usually is done and I’m called in to fix it. There are a lot of unnecessary antibiotics prescribed here.”
“Beware of patients who tell you they have a penicillin allergy,” said Dr. Isada. While 10% of hospitalized patients give a history of penicillin allergy, 90% of these people can actually tolerate the antibiotic. True IgE-mediated anaphylaxis to penicillin is present in only 1 or 2 per 10,000 treatment courses.
“The good news is that you can give most patients penicillin,” he said. “The bad news is that there is somebody that you can’t give it to. The best way to figure out who that might be is by taking a good history.” About 80% of patients with a true IgE allergic reaction at one time “can lose that sensitivity in 10 years.”
Also, details about reactions can be telling. While only Type 1 IgE reactions will lead to anaphylaxis, most hypersensitivities are not Type 1. Reactions include a rash that appears two weeks later, serum sickness, Stevens-Johnson, erythema multiforme, interstitial nephritis or a fixed drug eruption.
“If you are unable to determine if the patient has an IgE reaction by your H&P,” Dr. Isada said, penicillin skin tests can be “useful for intermediate probability situations.” This test can’t be performed on patients taking H1 or H2 blockers or SSRIs, he added. But if it is done correctly, the skin test has a negative predictive value of between 97% and 99%.
If you are still worried about a possible allergy but want to give penicillin, try a “graded challenge,” giving oral cephalosporins in slightly increasing doses to see what dose patients can tolerate. “One of the controversial areas about this graded challenge is that your risk of a problem is only about 2%,” Dr. Isada said, “but within that 2%, it’s still possible to get anaphylaxis.”
Necrotizing soft tissue infections
In some of the scariest infectious disease emergencies, hospitalists regularly misdiagnose and undertreat at first. Patients are typically young people in the emergency department complaining of significant pain when they have what looks like only a minor muscle injury.
“Pain out of proportion is an important clue to identifying necrotizing fasciitis” or another necrotizing soft tissue infection, explained James Pile, MD, a hospitalist and infectious disease subspecialist at Cleveland Clinic. Other key symptoms, like fever or loss of sensation, show up only later, and the disease can progress very rapidly. “Your index of suspicion needs to be high.”
The role of imaging is controversial, Dr. Pile noted, but CT with contrast can be particularly helpful in diagnosing the disease.
Dr. Flaherty reminded hospitalists that necrotizing fasciitis isn’t cellulitis, and cellulitis does not evolve into necrotizing fasciitis. When you’re unsure, he said, get a surgery consult immediately. Delayed surgery puts the patient at increased risk of a bad outcome, and mortality risks are already high.
“If a patient is concerned, you should be concerned too,” Dr. Flaherty said. Cognitive bias “thinking that patients must be exaggerating their pain “is a major cause of delayed diagnosis, he adds.
Timing of antibiotics
When it comes to giving antibiotics as perioperative prophylaxis, Dr. Flaherty had this advice: “Better late than never does not apply.”
“The data say you have to give prophylaxis in the window right before patients get operated on, ideally between 30 and 60 minutes before the incision is made,” he explained. “If you wait until after the incision is made, you might as well have not given it at all.”
Timing is also key when giving antibiotics for sepsis and hypotension, if you suspect encephalitis, or if the patient might have bacterial meningitis.
“If you think they have encephalitis, the timing is critical to the outcome, so the sooner the better,” Dr. Flaherty said. The same seems to be true with sepsis and hypotension. “Start broad-spectrum antibiotics now,” he added. “You can always back off later.”
With bacterial meningitis, Dr. Pile pointed out, “there has been a debate for a long time over whether an hour or two of delayed antibiotics makes a difference. But now, good evidence suggests that time
matters.” If you suspect bacterial meningitis, “the message is clear: Get blood cultures. Get the appropriate empiric antibiotics on board. And at that point, get the head CT “which is indicated in many but not all cases “and the lumbar puncture.”
Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.
C. diff update
AS OF NEXT JANUARY, patients will be able to look up each hospital’s rates of lab identified C. difficile infection (C. diff). Hospitals have begun reporting these data to the Centers for Medicare and Medicaid Services, and their rates will show up on the Hospital Compare Web site in 2014. Hospitals’ rate of C. diff infections may later be included in Medicare’s value-based purchasing program and affect payments.
A “confluence of factors” is making the disease worse in many hospitals, according to Thomas G. Fraser, MD, vice chair of the infectious disease department at Cleveland Clinic. Contributing factors include a hospital population of older and sicker patients, lots of proton pump inhibitors and fluoroquinolones, and a more virulent C. diff strain that has been circulating broadly.
There also seems to be a notable increase in the number of community-acquired cases. A study from the CDC published this June found that more than one-third of the growing population of patients with community-associated C. diff had not received antibiotics previously, but one-third had been taking proton pump inhibitors. About 40% also had some “low-level outpatient health care exposure,” like a doctor or dentist appointment.
“Very few of the risk factors are modifiable,” Dr. Fraser said. “Maybe you can think about acid suppression, about how you give antibiotics, and about cleaning our environment and handwashing.” But most hospitals have found that even with concerted efforts, they “can’t get it to zero.”
In addition to doing what they can for prevention, infectious disease experts offered some advice for treating C. diff diarrhea cases:
If you are using PCR assay, one test is enough to diagnose the condition. Diarrhea is considered three or more loose bowel movements in 24 hours.
Do not give antiperistaltics. This is a common mistake, Dr. Fraser said. “You don’t want to slow down the toxin” leaving the system.
The best first-line therapy is metronidazole, but oral vancomycin is preferred for seriously ill patients. Do not use these medications intravenously; they will not reach the stool that way.
If a patient is NPO, vancomycin enemas are an option, said Carlos M. Isada, MD, Cleveland Clinic’s ID vice chair. “One of the larger series of case reports suggests that surgery was rarely necessary when you used intracolonic vanco,” he said, even though it’s unclear whether intracolonic vancomycin reaches the right side.
Subtotal colectomy is an option for some severely ill patients with C. diff and ileus. “Those with severe disease should at least get a surgeon to look at them,” Dr. Fraser said. “If their white blood count keeps rising despite medical therapy, I am worried.”