Published in the December 2017 issue of Today’s Hospitalist
WHEN IT COMES TO skin and soft tissue infections (SSTIs), several randomized controlled trials published in the last 12 months have led to significant changes in treatment. That’s according to Jennifer Babik, MD, PhD, an infectious diseases doctor and associate professor at the University of California, San Francisco (UCSF).
During a presentation at UCSF’s management of the hospitalized patient conference this fall, Dr. Babik broke down those infections into two main categories: purulent, which includes uncomplicated abscesses and purulent cellulitis, and nonpurulent, including nonpurulent cellulitis and necrotizing skin and soft tissue infections. She also spelled out the best treatment options for each, incorporating the latest evidence.
What’s the best treatment option for uncomplicated abscesses: no antibiotics, trimethoprim sulfamethoxazole (TMP-SMX), doxycycline or amoxicillin/clavulanate?
Just a year or two ago, “no antibiotics” would have been the right answer, Dr. Babik pointed out. But two new studies—one in the March 3, 2016, issue of the New England Journal of Medicine (NEJM), the other in the June 29, 2017, NEJM—both showed that antibiotics were better for patients with small abscesses than incision and drainage alone.
“With clindamycin, you can have resistance. It’s really important to know your local susceptibility.”
~ Jennifer Babik, MD, PhD
University of California, San Francisco
“In the first study,” Dr. Babik pointed out, “TMP-SMX proved to be better than placebo with an absolute risk reduction of 7%, so the number needed to treat was 14.” The second study compared TMP-SMX to clindamycin and found the two drugs to be equivalent, while both were better than placebo. “Here the risk-reduction difference with antibiotics was even greater, at 13% or 14%, so the number needed to treat was only 7.” Together, the studies have changed the way ID providers manage uncomplicated abscesses by adding adjunctive antibiotics.
Both studies also confirmed that MRSA accounts for between 55% and 60% of purulent SSTIs, while MSSA causes 17%, Dr. Babik noted. “In thinking about empiric oral antibiotics, the key is to target S. aureus and include MRSA coverage.”
Drug options; purulent cellulitis
Several drugs do target known MRSA, each with pros and cons. With TMP-SMX, doctors should consider a five-to-10 day course, one double strength BID. As Dr. Babik pointed out, TMP-SMX is not usually considered a very good drug for Streptococcus, and “it has a lot of side effects, as you all know: hyperkalemia, acute kidney injury and hypersensitivity effects.” But it’s very good for S. aureus and, importantly, is more than 95% sensitive for UCSF’s local S. aureus isolates.
Doxycycline, which is also commonly used for this indication, likewise doesn’t have good coverage for Streptococcus. But while most S. aureus is still susceptible to it, the drug has less data backing it than TMP-SMX.
While clindamycin has very good activity against Streptococcus, it also carries a risk of C. diff—and “with clindamycin, you can have resistance. This is where it’s really important to know your local susceptibility.” In San Francisco, for example, only between 45% and 65% of MRSA is susceptible to clindamycin.
As for linezolid: “Occasionally, this will come up when you need an oral MRSA drug,” Dr. Babik said, “but it’s expensive and it has some pretty severe side effects,” including marrow suppression.
With known MSSA, on the other hand, “switch to cephalexin or dicloxacillin,” she said. “We’re more comfortable with cephalexin, so that would be my drug of choice.” Doctors could use any of the MRSA options for patients with a penicillin allergy. And although the Infectious Diseases Society of America (IDSA) recommends doing abscess cultures, “that’s not always done.”
For purulent cellulitis, “use an IV agent in patients with moderate to severe infection,” said Dr. Babik. The first choice is usually vancomcyin.
But what if patients have a vancomycin allergy? Her choice would be daptomycin, particularly when patients are sick and she’s worried they have S. aureus and may be bacteremic. While clindamycin covers Staphylococcus, “it is not a great agent for bacteremia because it’s a bacteriostatic, not bacteriocidal,” she said. Studies have found TMP-SMX to be inferior to alternative agents for serious Staphylococcal infections. And “I don’t think you need to add pip/tazo for gram-negative coverage.”
Nonpurulent cellulitis and blood cultures
Nonpurulent cellulitis is much more common, Dr. Babik pointed out, making up about 85% of cases. When you suspect nonpurulent cellulitis, should you get blood cultures?
“As an ID person, I love to culture everything at all times,” she said. “However, data say you don’t need blood cultures because they are positive in nonpurulent cellulitis in less than 5% of cases.” Even in cases identified, “it rarely changes management.” That said, doctors “need to use their clinical judgment in terms of what the person looks like.”
Patients with no or very low fever and a low white count don’t need to be cultured. As for which patients do per IDSA guidelines, those include high fever—”which the IDSA doesn’t de- fine; I’d say in the high 38s”—and significantly elevated white count; those who are immunocompromised, including malignancy; or those at risk for unusual organisms from an immersion injury or animal bite.
“I’d also add someone who has a risk of S. aureus bacteremia,” she said. “When in doubt, get blood cultures.”
What about skin biopsy or aspiration? According to Dr. Babik, the yield from needle aspiration has been “all over the map” in studies, from less than 5% to more than 40%, while the yield of skin biopsy “pretty much sits at about 20%.” Although she doesn’t commonly do skin biopsy, she considers it when patients fail to respond to what she considers appropriate treatment, especially those with severe infection or who are immunocompromised.
“I’d biopsy a patient who is neutropenic and has a
severe infection because I’d be worried about some-
thing unusual, like a fungal infection,” she said. She’d
also be concerned about a cellulitis mimicker. “When
nothing seems to add up and you’re worried some-
thing else may be going on, do a skin biopsy and get dermatology involved.”
Covering beta-hemolytic Streptococcus
Which antibiotic would Dr. Babik start these patients on? Her choice would be cefazolin.
“Vancomycin wouldn’t be incorrect, but I think it’s probably broader than you need,” she noted. “In our hospital, almost 100% of these patients get vancomycin, but cefazolin would be a better choice.”
Studies (including one in the January 2016 Open Forum Infectious Diseases) show that as much as 85% of nonpurulent SSTIs are caused by beta-hemolytic Streptococcus. More than 95% of these patients improve when given beta-lactams.
“S. aureus is found only 3% of the time, while nothing is identified in 12% of cases,” Dr. Babik said. “Almost all these patients got better on a beta-lactam without MRSA coverage.”
But doctors apparently aren’t getting that message. A study in the May 2014 issue of The Western Journal of Emergency Medicine looked at ED patients being treated for nonpurulent cellulitis. “In 2007, 56% of people got MRSA coverage,” she pointed out. “In 2010, it was 68%.”
In patients with moderate to severe infection, Dr. Babik opts for IV therapy, and “cefazolin is my go-to agent,” she noted.
“That’s partly because it gets MSSA as well, but also because it’s been well-studied, at least in its oral equivalent—cephalexin—for uncomplicated cellulitis.”
She does cover MRSA in patients with severe infection or “an infection in the setting of surgical site or injection drug use,” patients who are immunocompromised or have penetrating trauma, patients who have MRSA elsewhere, or those who are just not getting better. IV options for MRSA coverage include vancomycin, daptomycin or linezolid. PO options include clindamycin alone or a beta-lactam plus either doxycycline or TMP-SMX.
When to escalate
As for covering possible gram negatives, indications include severe infection—”you’re worried it might even be a necrotizing infection,” said Dr. Babik—or someone who’s neutropenic. Indications also include surgical site infections in the abdomen or axilla, patients with orbital cellulitis “because we know you can have gram negatives there,” or, again, patients not getting better on other therapies.
While many IV options cover gram negatives, “I’m usually doing pip/tazo, cefepime or ertapenem,” she said. Step-down oral options include amoxicillin clavulanate or a fluoroquinolone, among others.
With patients started on cefazolin who don’t get better, how long do you wait to escalate? A study in the Oct. 15, 2016, issue of Clinical Infectious Diseases looked at hospitalized patients with nonpurulent cellulitis, measuring both the cessation of spread and improved inflammation. Researchers also measured whether patients defervesced and if their white blood cell count went down by 20% vs. the day before.
By day 3 after starting antibiotics, 98% of patients had improved inflammation, and 85% had lower fever and white count. “It took people a couple of days and, interestingly, the exam got better before the fever and white count,” she pointed out. Researchers also looked at patients whose therapy was escalated within two days. That was “super common, but it wasn’t associated with any difference in outcome, so it was probably done prematurely.”
When doctors do jump the gun and escalate patients to, say, vancomycin, “then I get asked: What should patients be discharged on?” Dr. Babik said. Often, her answer is cephalexin “because I think the person probably didn’t need to be escalated,” but each case has to be considered individually.
As for how long to treat, IDSA guidelines say five days, as long as there is clinical improvement. But a study published in November 2016 by BMC Infectious Diseases found that only 20% of patients with uncomplicated SSTIs (purulent and nonpurulent) received treatment for less than 10 days.
“For anybody involved in stewardship at their hospital,” said Dr. Babik, “this is certainly an area where we can improve.”
Necrotizing soft tissue infection
Dr. Babik finds the term “necrotizing soft tissue infection” more inclusive than necrotizing fasciitis. It’s a rare diagnosis, with only 500 to 1,500 cases in the U.S. every year, “but it’s very severe with a mortality of 20% and amputation in the 16% range.”
Among cases, 75% have positive wound cultures, while 35% have positive blood cultures, “so you definitely want to get blood cultures.” Infections break down about 50/50 into two categories: polymicrobial, which is a mix of gram positives, gram negatives and anaerobes, and monomicrobial.
“Monomicrobial is usually group A Streptococcus, S. aureus or gram-negative rods,” Dr. Babik noted. “Only rarely do you see clostridium, vibrio or fungal infections.” With negative cultures, she added, “assume it could be polymicrobial.”
If you suspect necrotizing soft tissue infection, an early surgical evaluation is key. Unfortunately, only 15% of patients are diagnosed on admission. Common findings are tenderness, erythema, warmth and fever. Uncommon findings—at least early on—include bullae, necrosis, hypotension, crepitus, induration and fluctuance. Two other important clues, Dr. Babik added, are pain seemingly out of proportion to the presentation and pain outside the margins of inflammation.
Go big with antibiotics
The LRINEC (laboratory risk indicator for necrotizing fasciitis) score is composed of a CRP, which is often missing at diagnosis, as well as white count, hemoglobin, sodium, creatinine and glucose.
“You can look this up on any medical calculator,” she noted. But several studies—including one in the June 2017 issue of The Western Journal of Emergency Medicine—have found that the score’s positive predictive value is “pretty bad,” while its negative predictive value “is OK but not perfect. A low LRINEC score cannot rule out this diagnosis, and it shouldn’t replace your clinical suspicion.”
As for imaging, plain films vary in sensitivity, and only 25% will pick up on these cases. CT is “a fair amount better,” with sensitivity and specificity running between 80% and 95%. “You’re looking for things like fascial thickening or muscle involvement.” MRI is overly sensitive, with a specificity of only 60% or 70%, and it takes a long time.
“Early surgery is critical,” Dr. Babik pointed out. Survival is 93% with surgery within 24 hours and drops to 75% at 48 hours.
In terms of antibiotics, “you want to go pretty big upfront,” with vancomycin plus either pip/tazo or carbapenem plus clindamycin. Use clindamycin if “you’re concerned the person could have group A Streptococcus” because, in this instance, it circumvents the Eagle effect.
“Penicillin and other beta-lactams are less effective with group A Streptococcus when it is growing at a very high inoculum rate like in necrotizing fasciitis,” Dr. Babik pointed out. “Clindamycin gets around that.”
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
Treating animal bites
ANIMAL BITES don’t fit neatly into any one skin and soft tissue infection category, pointed out Jennifer Babik, MD, PhD, an infectious diseases specialist at the University of California, San Francisco (UCSF). She spoke at this fall’s management of the hospitalized patient conference there.
“Bites present as nonpurulent cellulitis in about a third of cases, purulent cellulitis in about half and an uncomplicated abscess in about 15%,” Dr. Babik said. The majority of bite infections are polymicrobial, with Pasteurella found in 75% of cat bites and 50% of dog bites, along with Streptococcus, Staphylococcus and anaerobes. (Eikenella is found in 20% of human bites.)
Patients should receive antibiotic prophylaxis if they are immunocompromised or have advanced liver disease or edema of the affected area. Prophylaxis is also indicated for moderate to severe injuries, particularly of the face or hand, or with deep penetrating wounds, such as cat bites. According to Dr. Babik, “your first-line choice is amoxicillin clavulanate for three to five days,” while moxifloxacin and doxycycline are good alternatives. Patients also need to make sure their tetanus is up to date.
As for established infection, ampicillin/sulbactam is “the IV drug of choice,” because it covers Streptococcus, Staphylococcus and gram negatives. Add MRSA coverage for purulent infections, while oral agents are the same as those for prophylaxis.
Dr. Babik also noted that a UCSF hospitalist recently shared a case with her of a patient with nonpurulent cellulitis whose blood cultures grew Capnocytophaga, a type of gram negative. “It turns out,” she said, “the patient was having his dog lick his wounds.”