Abscess Management: The Reformation of an Antibiotic Nihilist

Abscess management has evolved somewhat in the 14 years since my residency graduation. The point at which antibiotics are likely to be more helpful than harmful is not always easy to assess, and evidence based expert opinion has flip flopped impressively.

Based on current evidence, I would like to answer 3 big questions that every clinician may have when confronted with an abscess:

  1. Who needs antibiotics?
  2. Which abscesses need to be drained?
  3. How should abscesses be drained?

A Brief History:

Community Acquired MRSA has caused an increase in ED visits for skin and soft tissue abscesses since the turn of the century.  In 2003 after >1000 LA county prison inmates contracted skin and soft tissue abscesses after ‘spider bites’ extensive and expensive efforts were put into eradicating spiders from LA county prisons.  This did not change the rates of skin and soft tissue infections.  Despite the multiple patients presenting with ‘spider bites’ who are besmirching the good name of spiders everywhere we know that almost all ‘spider bite’ abscesses are due to CA-MRSA. CA-MRSA has also affected multiple professional sports teams, including the NFL.

Who Needs Antibiotics?

Emergency physicians have a front row seat to patients presenting with complications of indiscriminant use / overuse of antibiotics. We are beset upon on a daily basis by the worried well who have expectations of receiving antibiotics for a myriad of complaints where antibiotics will not help and may hurt.  We are attracted to evidence that supports fewer medication interventions, thereby decreasing the incidence of medication related adverse events.

The Evolving Recommendations for Antibiotics After Abscess Drainage:

2007 Hankin, Everett, Are antibiotics necessary after incision and drainage of a cutaneous abscess? Systematic review. Annals of EM, 2007.  PMID: 17577944 

  • 5 studies, 1 abstract, 30 year review.
  • No clarification of degree of overlying cellulitis in the studies reviewed.

2010 EBmedicine.net. Emergency Department Infections in the Era of Community-Acquired MRSA. October 2010.

  • Recommend empiric CA-MRSA coverage if:
  • Systemic signs and symptoms are present
  • Comorbidities (DM, immune system impairment)
  • Critical anatomic location (face, genitourinary)
  • Surrounding cellulitis
  • Large size (No specific size is mentioned)
  • 5cm abscess size cutoff?
    • The 5cm cutoff /breakpoint for antibiotic administration is an interesting choice. There is variation in the literature of whether to measure the diameter of total infected site, versus clinically apparent area of fluctuance, versus confirmed cavity size as measured by bedside ultrasound.

Lee et al. Pediatric Infectious Disease Journal 2004.  PMID: 14872177

  • 69 patients. Prospective cohort study.
  • I and D alone did not resolve abscesses larger than 5cm.
  • All patients received antibiotics.(mostly cephalexin, or amoxicillin)
  • 58 patients managed as outpatients with CA-MRSA abscesses.
  • No differences in outcome in the patients whose antibiotics were changed to CA-MRSA specific coverage.
  • Terminology: Abscess size vs infected site diameter. This study uses ‘infected site diameter’.

Any time we are using our measurements to make clinically important decisions we need to take into account the accuracy of the measurements we are obtaining. As it turns out, we are not as good at obtaining measurements as we would expect:

Measurement – Size matters: how accurate is clinical estimation of traumatic wound size? Injury 2014  PMID: 22592151

  • 50 providers evaluated wounds, asked about management, and estimate of wound length.
  • There was significant inter-provider variability.
  • Men consistently over estimated wound length. Women consistently under estimated wound length.
  • If we are not good at estimating laceration size, abscess size estimation is unlikely to be any different.

New Evidence Since 2010:

Talan D et al. Trimethoprim-sufamethoxazole versus Placebo for Uncomplicated Skin Abscess. NEJM 2016   PMID: 26962903 

  • 1247 Patients. I and D alone versus I and D plus Bactrim.
  • Primary outcome was cure rate at 7-14 days. (No fever, decrease / no increase in size of infected area, decreased discomfort)
  • Ultrasound was used in all abscesses where fluctuance wasn’t obvious. Use of US allowed for accurate measurement of abscess.
  • Average area of erythema: 6.5-7.0 x 5 cm
  • 20% of patients had >75 sq. cm of erythema (8.7cm diameter)
  • Cure rate: 80.5% (Bactrim) vs 73.6% (placebo)
  • Secondary outcomes:
    • Subsequent surgical drainage: 3.4% (Bactrim) vs 8.6% (placebo)
    • Skin infections at new sites: 3.1% (Bactrim) vs 10.3% (placebo)
    • Infections in household members: 1.7% (Bactrim) vs 4.1% (placebo)
    • 1 episode of hypersensitivity reaction, fever, hepatitis, thrombocytopenia. Spontaneously resolved.
    • No episodes of C-diff.
  • Conclusion:“In MRSA prevalent areas, treatment with Bactrim in addition to I and D resulted in higher cure rate”.

Daum, R et al. A Placebo-Controlled Trial of Antibiotics for Smaller Skin Abscesses. NEJM 2017  PMID: 28657870

  • 786 patients. I and D alone vs I and D plus either clindamycin or Bactrim.
  • Primary outcome was cure rate at 7-10 days.
  • Excluded patients with temp >38.5C, or those meeting SIRS criteria.
  • Average area of surrounding erythema 26 sq. cm. (5×5).
  • Cure rate: 83.1% (clinda), 81.7% (Bactrim) vs 68.9% (placebo)
  • Treatment related adverse events: 21.9% (Clinda) 11.1%(Bactrim) vs 12.5% (placebo)
    • NO C-diff reported
  • All participants had moderate sized areas of overlying cellulitis.
  • Conclusion: “the increased clinical cure rate must be weighed against the potential risks associated with increased antibiotic use”.

Talan et al. Subgroup analysis of Antibiotic Treatment for Skin Abscesses. Annals of EM 2018. PMID: 28987525 

  • Subgroup analysis of patients in Talan 2016 NEJM study.
  • Found consistent treatment effect for erythema / infected site size >5cm and <5cm
  • Biggest treatment effect through 42-56 days seen in those with a history of MRSA (76.7% vs 56.8% (22.9%) ) and fever (77.6 % vs 60.6% (16.9%) )

Gottlieb et al. Systemic Antibiotics for the Treatment of Skin and Soft Tissue Abscesses: A Systemic Review and Meta-analysis. Annals of EM 2018.  PMID: 29530658  

Reviewed on REBEL EM April 2018

  • 2406 patients in 4 studies. (heavily influenced by Talan and Daum papers)
  • Treatment failure rate: 7.7% (antibiotic) vs 16.1% (placebo) Calculated risk difference 7.4%
    • NNT: 13.5
  • New lesions at different sites: 6.2% vs 15.3%,
  • Absolute difference 9.1%, calculated risk difference 10%
    • NNT 11
  • Diarrhea: 11.8% vs 11.2% (calculated risk difference 0.8%)
    • No cases of C-diff.

Overall, the new evidence from high quality studies since 2010 suggests a liberal antibiotic strategy would likely improve already high cure rates, may decrease MRSA recidivism, and surprisingly could do so without generating more antibiotic associated diarrhea.

Rare Antibiotic Side Effects:

We routinely caution patients against the common side effects of diarrhea, candidiasis, nausea, and also a potential increase in resistant strains of bacteria. What about the rare, and sometimes life threatening antibiotic associated skin conditions such as Stevens Johnson syndrome and toxic epidermal necrolysis? The true rates of these rare conditions are hard to calculate, 2 old studies suggest the following rates:

Roujeau J, et al. Medication use and the risk of Stevens Johnson syndrome and toxic epidermal necrolysis. NEJM 1995. PMID: 7477195  

  • Population case control study to estimate incidence of SJS / TEN following exposure to several agents known to be associated with these conditions.
  • Highest association: Sulphonamides, (most commonly Bactrim)
  • 4.5 cases per 1 million users per week

Chan H, et al. The incidence of erythema multiforme, Stevens-Johnson syndrome, and Toxic Epidermal Necrolysis. A population based study with particular reference to drugs uses among outpatients. PMID: 2404462 

  • 14 year observational study. Single health system in Seattle.
  • 61 suspected cases, 16 cases attributed to drugs given before admission.
  • 3/100,000 person years for Bactrim (by comparison, Nitrofurantoin was 7/100,000 person years)

Who to use an Antibiotic On?

  • Cellulitis that you would treat if there was no abscess (including infection site size <5cm in diameter)
  • Unreliable patient (poor health literacy, poor access to follow up)
  • Big abscess (larger that 5cm)
  • Immunocompromized (DM, steroids, elderly, significant systemic disease burden)
  • MRSA recidivist. MRSA in the house, any history of previous abscess

Who to Drain?

  • Some abscesses are obvious, and the use of ultrasound will be a waste of time.
  • In abscesses with a lot of skin thickening, or in cellulitis without obvious fluctuance it is not always obvious which cellulitis has an abscess lurking beneath.
  • Use Ultrasound if you are not sure:

Babic et al. In patients presenting to the emergency department with skin and soft tissue infections what is the diagnostic accuracy of point-of-care ultrasonography for the diagnosis of abscess compared to the current standard of care? A systematic review and meta-analysis. BMJ open 2017.  PMID: 28073795

  • Ultrasound sensitivity: 96.2%
  • Ultrasound specificity: 82.9%
  • If you wouldn’t bet $5 on whether a drainable abscess is present, use an ultrasound.

How to Drain?

Historically we have always packed abscesses. This has resulted in scheduled follow up visits for packing change / removal. A new (2010) technique of loop vessel drainage can potentially reduce / avoid follow up visits.

 Loop Vessel Technique

Am J Emerg Med, 2018. Comparison of the loop technique with incision and drainage for soft tissue abscesses: A systematic review and meta-analysis. PMID: 28917436 

  • 4 Studies, 470 patients.
  • Comparison of CID (conventional incision and drainage with packing) and LVT (Loop vessel technique)
  • Failure rate: CID 9.8% vs LVT 4.1%

The acid test of any new intervention is whether the patients like it. Some of my happiest patients are those with recurrent shooter’s abscesses who experience their first loop vessel drainage, and are extremely happy to not have to deal with a packed abscess.


  1. Incise at one edge of abscess
  2. Insert mosquito forceps / yankauer suction catheter
  3. Break down loculations
  4. Tent the skin opposite the first incision
  5. Cut down to the catheter / mosquito forceps
  6. Drag Loop Vessel back through the abscess
  7. (consider irrigation)
  8. Make 5 throws on the first tie
  9. Tie tight, but ensure no skin tension
  10. Remove vessel loop in 5-7 days

For style points, try not to spill a drop of pus. As healthcare providers a big spill of pus is fun, but not so for our patients.

Take Home Points:

  • There is room for a safe increase in antibiotic use
  • There does not need to be reckless over-use of antibiotics
  • Use ultrasound with any abscess you are unsure of
  • Use a loop vessel rather than packing

Post Peer Reviewed By: Salim R. Rezaie, MD (Twitter: @srrezaie)

Cite this article as: Rob Bryant, "Abscess Management: The Reformation of an Antibiotic Nihilist", REBEL EM blog, May 12, 2018. Available at: https://rebelem.com/abscess-management-the-reformation-of-an-antibiotic-nihilist/.

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share via Email

Want to support rebelem?