The clinical milieu of the Emergency Department is far from straightforward. The cognitive grime in which we function is rarely conducive to the pristine distinctions drawn by Evidence-Based Medicine (EBM). As such we are often asked to make dichotomous decisions based off a single study, using answers to questions we never intended on asking.
Until recently the management of simple cutaneous abscesses was as straightforward a complaint as there was in Emergency Medicine. Incision and drainage was all that was required. Packing, irrigation and antibiotics had all but been discredited by those in the evidence based know. At least that was until someone went and studied it. A recent trial by Talan et al published in the NEJM examined the utility of enteral antibiotic therapy when added to incision and drainage (I&D) for the management of simple cutaneous abscesses (1). The authors randomized 1265 patients presenting to the Emergency Department with cutaneous abscesses to either I&D and placebo or I&D with the addition of a 7-day course of trimethoprim-sulfamethoxazole. The authors found that patients randomized to the 7-day course of antibiotics had a clinical cure rate of 80.5% vs. only 73.6% in the patients who received I&D alone. The use of trimethoprim-sulfamethoxazole was also found to be superior in a number of secondary endpoints, resulting in lower rates of subsequent surgical drainage procedures (3.4% vs. 8.6%), skin infections at new sites (3.1% vs. 10.3%), and infections in household members (1.7% vs. 4.1%) 7 to 14 days after the treatment period. These improvements in cure rates came at the price of a small increase in gastrointestinal distress, with a 6.5% absolute increase in the rate of GI complaints in patients randomized to receive trimethoprim-sulfamethoxazole.
This was a methodologically rigorous trial, with statistically significant results. Likely its findings will lead to a global shift in practice towards prescribing antibiotics for all simple cutaneous abscesses. And yet despite the strength of these findings, is such a change in practice justifiable?
Talan et al conducted a pragmatic trial, in which they enrolled all patients older than 12 years old presenting to the Emergency Department with surgically confirmed abscesses larger than 2 cm in diameter conducive to outpatient treatment. The authors limited their exclusion criteria to indwelling devices, suspected osteomyelitis or septic arthritis, diabetic foot ulcers, ischemic ulcers, mammalian bites, wound containing organic foreign bodies, infection of another organ system, perirectal, perineal or paronychial location, intravenous drug use within previous month with fever, underlying skin condition, residence in a long-term care facility, incarceration, or known immunodeficiency. This practical design ensured a heterogeneous cohort of abscesses. A large portion of which had significant concomitant cellulitis. The median area of surrounding cellulitis was 20 cm2 , with approximately 20% of the cohort presenting with a surrounding cellulitic area of greater than 75 cm2 . 11% of the patients had diabetes and 18% had a fever in the days leading up to their presentation to the Emergency Department. Essentially this was not a cohort made up entirely of simple cutaneous abscess. It contained a large quantity of infections that would likely benefit from antibiotics. Despite this, the majority of patients experienced resolution of their abscess with I&D alone. Of the patients who were deemed a clinical failure, very few required further medical therapy, their abscesses ultimately resolved with no further treatment. There was only a 5.6% increase in the number of patients who required addition surgical drainage and a 7% increase in the number of patients started on addition al antibiotics at the follow-up visit between 7-21 days.
The pragmatic design of this study makes for an easy translation to our general Emergency Department population. Unfortunately this very same pragmatism limits a finer, more granular interpretation of the data. In a population that includes abscesses of various shapes and sizes, with varying degrees of surrounding cellulitis, a seven-day course of trimethoprim-sulfamethoxazole provides a moderate degree of clinical value above I&D alone. But it can also be said that the vast majority of patients presenting with cutaneous abscesses will fare equally as well with simple surgical drainage. The question remains, are the patients that failed I&D alone clinically predictable? Meaning were these the patients who had significant surrounding cellulitis, systemic symptoms (fever, malaise) or baseline immune-compromised conditions (diabetes, etc.)? Furthermore this trial fails to address the more subtle unattended consequences of empirically treating all simple cutaneous abscesses with trimethoprim-sulfamethoxazole. This trial was severely underpowered to assess the rate of rare reactions such as Steven Johnson’s syndrome (SJS) associated with the use of trimethoprim-sulfamethoxazole (2). Nor can it evaluate the effects such a change in practice will have on current antibiotic resistance.
Talan et al clearly demonstrate that the majority of patients with simple cutaneous abscesses will improve with I&D alone. The ones that do fail are likely clinically predictable upon presentation. Most will resolve with no further management, with a small minority requiring further surgical drainage or the addition of antibiotics. For the well appearing patient presenting with a simple cutaneous abscesses with minimal cellulitis, a trial of I&D alone seems reasonable.
Clinical medicine is far more complex than the yes/no dichotomy mandated by frequentist methodology. And while the Talan et al trial has provided us with clear evidence that some abscesses will benefit from antibiotics, it does not demonstrate that all abscess do so.
- Talan DA et al. Trimethoprim-Sulfamethoxazole versus Placebo for Uncomplicated Skin Abscess. NEJM 2016; 374 (9): 823 – 3
- Chan HL, Stern RS, Arndt KA, et al. The incidence of erythema multiforme, Stevens-Johnson syndrome, and toxic epidermal necrolysis: a population-based study with particular reference to reactions caused by drugs among outpatients. Arch Dermatol 1990; 126: 43-7.