October 29th, 2016

Are Antibiotics Useful for Small Skin Abscesses? Now There’s an Answer

A "few" more steps is quite the understatement.

A “few” more steps is quite the understatement. And who chose that carpet?

Let’s start with the clinical controversy, one that’s been bouncing around Emergency Rooms, outpatient practices, postgraduate courses, and medical journals for years. Specifically, are antibiotics helpful for skin abscesses that are adequately drained?

It’s still debated since of course most patients with this annoying problem will get better on their own provided the drainage is adequate. What do the antibiotics add, if anything?

Now, thanks to a superb clinical trial presented this past week in New Orleans at the annual IDWeek Meeting, we have some answers.

And though the study is important, there’s a good chance you don’t know about it, even if you attended IDWeek. First, just like the last time this group presented results of a different trial on at a national meeting, the abstract presentation took place in front of a only a handful of meeting participants.

Second, even if you wanted to come see the presentation, you might have been stymied by the vast travel distances inside the New Orleans Convention Center, which must be one of the longest indoor spaces on the planet. Safe to say that no one who attended this meeting finished a day with fewer than 10,000 steps, unless they brought along their hoverboards or Segways. Sometimes it felt like I was walking half the way to Baton Rouge.

(For the record, that first study ended up being a major paper in the esteemed New England Journal of Medicine. Justice!)

Anyway, here are some of the details of this new study:

  • Adults and children with small skin abscesses were eligible provided they weren’t immunocompromised, diabetic, or systemically ill.
  • They all underwent abscess drainage and cultures.
  • Participants were randomized 1:1:1 to receive TMP/SMX, or clindamycin, or placebo for ten days. (Glad they followed the Golden Rule.)
  • 786 subjects enrolled (505 adults and 281 children) at several sites around the country.
  • 67% had positive cultures for Staph aureus; 74% of these isolates were MRSA.
  • At a 10-day post-therapy visit, cure of infection was seen in 83.1% and 81.7% of clindamycin and TMP/SMX treated patients respectively — both significantly better than placebo (68.9%).
  • TMP/SMX was better tolerated than clindamycin, with adverse event rates of 11% (TMP/SMX) and 22% (clindamycin). Placebo was 12.5% (guess that’s the “nocebo” effect). There were no cases of C diff or severe rashes.
  • Clindamycin-treated patients had fewer relapses than those receiving TMP/SMX.

The droll presenter, Dr. Robert Daum, and the senior author, Dr. Chip Chambers, made a couple of other interesting observations. First, if the culture grew a Staph aureus resistant to clindamycin, responses were reduced. The good news is that this resistance was observed in only around 5% of isolates, a much lower percentage than typically seen in Staph aureus from hospitalized patients.

Second, if cultures were negative for staph, response rates were similar between all three treatment arms — meaning you could potentially stop antibiotics in select patients based on culture results (though important caveat, this strategy wasn’t tested).

The results of this study add to a growing body of evidence that antibiotics do in fact improve outcomes even for drained skin abscesses, swinging the pendulum further back in the direction of using them for this indication. Whether you and your patient think that this net 10-15% increase chance of cure is worth the risks of antibiotics will probably depend on the case — but at least now we have the data to make an informed decision.

And boy, do my feet hurt.

24 Responses to “Are Antibiotics Useful for Small Skin Abscesses? Now There’s an Answer”

  1. Great post, thanks Paul!

  2. Jeff Dickey says:

    Don’t even think about trying it in high heels.

  3. Fares Masannat says:

    I rarely ever see a patient in the ER with a skin abscess, so I think the major audience for this study should the the ER/ urgent care provider. I hope there’s a way we can get them to read those results.
    If a great article is written about UTIs, I’ll never read it if gets published in a urology journal.

    Just like if there was a good restaurant at an RSA Charge conference I won’t know because I’m not willing to walk that far (ID week 2016 insider joke)

  4. Loretta S says:

    Important stuff, Paul. Thanks! Of course, I now have whiplash: “Give Abx AND drainage!”. Then, “Drainage alone is usually enough!”. Now, “Rate of cure is improved if we give Abx AND drainage!”.

    Does this stat frighten anyone else? “67% had positive cultures for Staph aureus; 74% of these isolates were MRSA.” Paul, I remember you writing that outpatient MRSA infections used to be extremely rare.

    • Paul Sax says:

      Fortunately rates of community (and hospital) aquired MRSA appear to be declining in many settings (including here in Boston). We’re definitely lower than this 74%, closer to 50%.

      The other bizarre thing that’s happening is that a higher proportion are now penicillin sensitive!


  5. michèle halpern says:

    About the greater number of relapses in the TMP/SMX arm: I wonder what dose was used. Was it wt based in adults? 1 DS bid, though ok for UTI’s, may not be enough for a skin infection, if based on using 5-10 mg/kg total daily dose of TMP. Higher doses may however cause greater GI or heme or renal/electrolytes side effects.

    • Paul Sax says:

      Clindamycin was dosed as two 150 mg tablets three times daily. TMP-SMX was dosed at two 80/400mg tablets twice daily.


      • michele halpern says:

        These doses are 50% lower than the recommended by IDSA (Practice Guidelines for the Diagnosis and Management of Skin and Soft Tissue Infections: 2014 Update by the Infectious Diseases Society of America), i.e clinda 300 (at least) mg tid or TMP/SMX 160/800 1 tab bid (2 bid for BMI>40).
        I am not saying that it explains the lack of full clinical response, and agree that higher doses may cause more side effects, but it seems strange to do things half way.

  6. Jesper says:

    Hi, sorry to discredit any good intentions from Paul, but if you read the original article that Paul was referring you will notice that it was a superiority trail of clindamycin vs Tmp-SMX, no placebo group exists !
    So either Paul deliberately tried to mislead, or he mixed up his references. Either way please let this be a lesson in critical reading !

    • Paul Sax says:

      Jesper, to be fair (to me), nowhere did I say that the prior study was one with a placebo arm. Thank you very much for reading this blog, and let the lesson in critical reading stand!

  7. Federico says:

    Dear Dr. Sax:

    Acording to “The Golden Rule” (which is priceless!) i wonder why they prescribe 10 days of ATB therapy instead of 7 that is the standard of care of uncomplicated skin abscess?
    Anyways, many thanks for posting this study! With all the buzz going around about this issue i was about to change my practice in order to aply Rational Use of ATB (thankfully i did not!).

    And about CA-Staph aureus, we are also seeing a higher proportion of MSSA. Maybe Staph aureus can also get Alzheimer’s! 🙂

    • Loretta S says:

      Very interesting about the trend toward increasing penicillin sensitivity of S. aureus that you and Paul noted. Maybe they can get Alzheimer’s. Or maybe the bacteria are up to something more sinister. You never know with those wily bacteria!

  8. Jon Blum says:

    I missed that abscess presentation. I’m still not sure I want to use antibiotics, and wonder if there is a way to predict who would benefit. For example, I would consider a 5 cm abscess “small” only if it’s not on me, and I wonder if smaller lesions would benefit less from treatment. At my facility, the last time we looked (in the mid-2000s, at the height of the USA 300 epidemic), only about 1/4 of the boils cultured in a nearby urgent care center were MRSA. TMP/SMZ and clinda are the two antibiotics most likely to cause an ED visit for an adverse reaction (http://cid.oxfordjournals.org/content/47/6/735.full) so I am a little surprised they did not see more problems with them. In the NEJM study, there were no serious adverse events, but 8-9% of the patients discontinued the drug due to an AE – not much less than the % of patients who benefited from treatment in the new study. (The older study had no placebo arm, so the real rate of problems may have been less.)

    In our small surveillance study a decade ago, the ED doc who cultured those boils predicted before the cultures returned whether the patient would have MRSA or not. He was correct an astounding 97% of the time. He later told me he used four criteria: recurrence, household member with a lesion, location (MRSA is groinotropic), and thick gooey pus. I’d love to see that replicated.

    • chris thomas says:

      I might add pain to the list of clinical indicators. One study I recall, demonstrated that MRSA elicited a more painful local reaction than other skin pathogens. I was always struck by the fact that these patients seemed to complain more strongly about the discomfort of these lesions.

      • Jennifer Beeckman says:

        As a provider and a once-upon-a-time MRSA sufferer, I can attest to the pain of these small boils. My first one was just at my hairline in front of my ear and the pain into my ear was incredible!
        After 2 years of going back and forth with these boils (and none of us had ever had boils before)- re-infections in myself, my husband, my toddler and then my baby- I had finally had enough. While ID treated my baby, I treated the rest of the family at the same time and used mupirocin in our noses as well ( I and my husband had cultured positive there) and we FINALLY got rid of it. And we are culture negative to this day and nary a boil between us.
        As a result, I totally believe in concurrent household treatment and the usefulness of antibiotics in abcesses.

  9. Dany says:

    Thanx for the publication of this trial, although I am not sure, and actually don t hope , this trial ” will swing the pendulum back to AB treatment besides drainage ” for skin abscesses in Europe in otherwise healthy people…

    If I Read the numbers well, we have to treat eight people to help one more patient as all the others would have had spontaneous recovery. What about the number needed to harm? It says ” severe skin reactions ” but less severe reactions they also happen in about one in eight in TMX/SMZ users and are nevertheless a big nuisance. And what about ten days of clindamycin that will alter intestinL flora for up to two years after exposure…

    And last but not least, what about the risk of creating superrbugs that will only be sensitive to iv AB?

    No, not for my own patients, not even for myself will I use a ten day ABcourse after successfull drainage. I care about our bacterial ecosystem, and I am happy that we have a lower than 20% rate of MRSA in many parts of Europe still.


    • Loretta S says:

      Great points. Anecdotally, it is usually around Day 10 of TMP/SMX that patients will call me, complaining of an all-over body rash (that call makes my stomach tighten). Fortunately, each time the rash has turned out to be the typical maddeningly-itchy but ultimately-resolving rash all of us have seen, but still… It has made me a bit leery of prescribing TMP/SMX for > 7 days.

  10. Joseph Cooper, MD says:

    New Orleans Convention Center was poorly designed and I agree much too far to traverse. Good post otherwise.

  11. Beyond Liberal says:

    I’d been troubled by skin abscesses most of my life. What finally stopped them? Thoroughly washing them out after drainage with Hibiclens followed by triple antibiotic ointment. They did the trick and haven’t had one since; it’s been 4 years. In my experience, systemic antibiotics aren’t really necessary and, as we are well aware, set patients up for resistance. We all know how compliant patients can be with a course of antibiotics. (snark meter off…)

  12. Adan Atriham says:

    Apologies if I missed it, but can someone explained what “cure” means in this study. What is less pain? Less missed school/work days? Repeat I&D? Avoiding admission for IV abxs?

    All of these can be thought as cure, however, they are not all the same. Two days less of pain is not an outcome I am particularly interested in and if I have to give abxs to 8-10 patients to save 2 days of pain for 1 patient, I am not reaching for my prescription pad.

  13. Scott Conrad Davidson says:

    I agree with the comment concerning the definition of cure.

    Is there an argument to do nothing because nearly 70% of patients achieve cure, and treat the failures the same way as those who did not achieve cure with SMX/TMP?

  14. Mitesh Desai says:

    In the era of genomic and proteomic diagnostics, is it gauche to suggest that perhaps the old gram stain test might be a way to decide if patients should start on bactrim / clinda? >10% rate of AEs with abx is not negligible…

HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.