In this series we are going to pick apart the arguments and evidence surrounding a common but controversial topic in Emergency Medicine. The goal is to take a hard look at the practices we are handing down to future generations of ED docs and see what is based on opinion and what is based on evidence.
Thrown for a Loop
Is a full incision down the length of an abscess the best way to I&D?
Abscesses used to be so simple. You’d poke it and if it felt right then you’d slice it, squeeze it, drain it, pack it, and discharge it. Well, a lot of things used to be simple, but in this modern era where even a coffee is no longer just small or large, it’s tall, grande, or venti. And just deciding what kind of milk to buy in the grocery store has become an agonizing decision (can milk made from almonds really be called milk?). Abscesses have gotten a little tricky. We’ve already discussed in our previous post that packing an abscess with gauze is likely just causing a bunch of unnecessary grief to all parties involved. Now the question is, should we really be slicing abscesses open or should we opt for the loop drainage technique?
Arguments for the loop drainage technique:
- Gaszynski R et al. ANZ J Surg. Jan 2018.
- Özturan IU et al. Am J Emerg Med. Jun 2017.
- Gottlieb M and Peksa G. Am J Emerg Med. Jan 2018.
- McNamara WF et al. J Pediatr Surg. Mar 2011.
- Aprahamian CJ et al. J Pediatr Surg. Sep 2017.
- Tsoraides SS et al. J Pediatr Surg. Mar 2010.
- Ladde JG et al. Am J Emerg Med. Feb 2015.
- Thompson DO. J Emerg Med. Aug 2014.
The loop drainage technique (LDT) may be an unfamiliar concept to many emergency physicians. As you can see in the image below, two small incisions (black and red lines) are made at the edges of the area of fluctuance, forceps are used to break up loculations, and then a vessel loop (blue) is pulled through the holes and tied to itself outside the patient (making a ring).
Now I know what you’re thinking, what in the world is a vessel loop and where can I find one? A vessel loop is a soft silicone string used during some surgical procedures. So, you can usually get them from the OR supply, but don’t worry about it. A study by Thompson et al shows that the cuff of a sterile glove can be used instead of a vessel loop.
The theoretical benefits of performing this procedure as opposed to conventional incision and drainage (CID) are the following:
- Less pain: CID is rated as the 2nd most painful procedure performed in the emergency department (Singer et al, 1999). LDT has the potential to be less painful due to the smaller incision size and lack of packing.
- Better cosmesis: CID requires a large incision in order to be effective. The incisions in LDT are each typically less than 5mm.
Are these benefits just theoretical, or are they fact? Are the cure rates the same between these two techniques or do these potential benefits come at a cost? Let’s take a look at the evidence.
The first paper studying this technique in cutaneous abscesses is published by Tsoraides et al in the Journal of Pediatric Surgery in 2010. They retrospectively examined the outcomes of 110 pediatric patients with cutaneous, non-pilonidal abscesses who underwent LDT by pediatric surgeons in the OR. There was no comparison group. They found that of the 110 patients, 6 required reoperation (5.4%); four underwent repeat LDT and two underwent CID after the initial LDT procedure. There was no discussion regarding the rationale for change in technique on reoperation. All 110 patients studied completely healed, and the authors concluded that LDT is a feasible and safe technique.
Aprahamian et al expanded on this initial study and published an updated retrospective of 576 patients. All patients received LDT, there was no control group. Only 24 (4.2%) patients had unplanned re-operations.
There have been several other retrospective studies examining the efficacy and safety of LDT in comparison to CID. McNamara et al retrospectively looked at 219 pediatric patients undergoing abscess management by pediatric surgeons. In this group of 134 CID and 85 LDT performed, they found that 4 patients in the CID group versus 0 patients in the LDT group required repeat intervention to obtain clinical cure. As an aside they noted that the LDT incisions appeared to have a better cosmetic outcome in the immediate follow-up period.
Ladde et al showed a significant failure rate of CID compared to LDT in pediatric patients. In 143 cases 52 were treated with LDT and only 1.4 % failed treatment versus 10.5 % in the CID group.
Most of the literature involves the pediatric population. However, a recent study by Gaszynski et al examined the use of LDT in the adult population and found that LDT was better tolerated than CID. The study was a retrospective review of prospectively gathered data on 63 patients in a rural Australian hospital. There were 4 physicians involved in the study, all of them surgeons. Three of them were instructed to perform CID on all patients presenting with cutaneous abscess. One of them performed LDT on all patients presenting with cutaneous abscess. If adequate anesthesia was not able to be obtained in the ED, drainage was performed in the operating room under general anesthesia. The CID group required general anesthesia 100% of the time. The LDT group required general anesthesia 70% of the time.
Now, clearly, there is a distinct difference between the practice pattern at this particular Australian hospital and the hospitals here in the US. Although this may be a small study with some obvious flaws, the fact remains that 30% of the LDT group in this study was able to be managed in the ED without need for general anesthesia, which indicates that LDT is less painful in comparison to CID.
So far, the study by Ozturan et. al is the only prospective randomized controlled trial evaluating LDT versus CID in adults managed by ED physicians. There were 3 study investigators who were trained in LDT prior to the study enrollment period. There were 23 patients in each group of adults ≥18 years old with cutaneous abscesses presenting to the ED. Immunocompromised patients and patients in whom the abscess could not be identified by ultrasound were excluded. The mean abscess diameter in LDT group was 3.2 vs. 3.0 in CID group. Interestingly, none of the patients included were given empiric antibiotics before or after the procedure. 13% of the patients in the LDT group had diabetes vs. 0% in the CID group. The primary outcome was change in abscess and cellulitis diameter at 7 days. Secondary outcomes were pain score, procedure duration, patient satisfaction, need for antibiotics, need for repeat procedure.
It is important to note that none of the outcomes in the Ozturan study reached statistical significance. However, given that there were only 23 patients in each group the study was obviously underpowered to detect a meaningful difference. When looking at the numbers themselves, though, the need for antibiotics and repeat procedures was almost twice as high in the CID group when compared to LDT. This may not have statistical significance but certainly seems clinically significant. What we need is a study well powered to show a significant difference.
Unfortunately, such a high powered trial does not exist. Gottlieb et al performed a systematic review and meta-analysis of the studies by Ozturan, McNamara, Gaszynski and Ladde. They found that CID resulted in treatment failure 9.43% (25/265) of the time, whereas LDT failed 4.10% (8/195) of the time. Other than the study by Ozturan, all of the trials included in this meta-analysis were retrospective which limits the strength of the conclusions that can be drawn from this meta-analysis. However, so far, there have not been any studies, retrospective or otherwise that have demonstrated any statistically significant benefit of CID over LDT.
Arguments for conventional incision and drainage:
Didn’t I just say there were no studies that demonstrate benefit of CID over LDT? Yes, but as we demonstrated above, we also don’t have great studies to show a benefit of LDT over CID in adult patients presenting to the ED. Most of the trials on this subject are performed by pediatric surgeons on pediatric patients. Thus, if you are an emergency physician who has never attempted this procedure before and have not received specific training in this technique it is impossible to know if you would achieve the same results as those studies. Now, if there were a study showing that ED physicians who did not receive formal training in this technique were equally skilled at this procedure compared to pediatric surgeons, these studies would be more relevant.
Schmitz et al compiles survey data of 350 ED physicians and other providers to describe the most common I&D techniques used. Linear incision was the most common technique (85%) and elliptical incisions were the next most frequent. LDT was not mentioned in the study suggesting ED physicians are currently not using this technique and are not skilled in performing LDT.
So, looking through all the data in support of LDT we find only one study in adult patients with ED physicians performing the abscess drainage. Ozturan et al was an underpowered study and did not show any difference between CID and LDT. It seems premature to change techniques with this lack of data. In other words, if it ain’t broke, don’t fix it.
There also are no studies that show improved cosmesis or discomfort during the procedure, even though proponents of the technique argue these points as well. Right now this is a good hypothesis, and logically makes sense, but just hasn’t been shown with data.
- Extended Analogy: otherwise known as indication creep. “If it works for kids in an OR it will work in adults in an ED. I mean, adults are just large kids.”
- We have pretty good data for one specific population of patients and one specific method of performing the technique. It is easy to assume that it could be applied to a broader patient cohort but so far this hasn’t been shown by data. This is why we need large trials with diverse patient populations to have generalizable data.
- Hasty Generalization: “Ozturan was a small study but almost showed statistical significance, if more patients were enrolled it would show the real benefits.”
- Small sample sizes can cause statistical anomalies that are less likely in large sample sizes. It is dangerous to assume a positive or negative result seen in a small sample size will grow with the sample. Typically, there will be a regression to the mean.
- Example: A coin flipped 10 times in a row may end up heads 20% of the time, a coin flipped 1000 times or 10,000 times will land heads up much closer to 50% of the time.
Take Home Points:
- There is decent evidence to support LDT as the gold standard treatment for pediatric abscess by pediatric surgeons.
- We need more studies in adults patients and performed by emergency physicians before we can draw conclusions on the benefit of LDT over CID treatment in the ED.
- In theory LDT should have better cosmesis and less pain than CID but no studies to date have looked at these outcomes specifically.
In summary, there is only weak evidence suggesting that LDT is superior to CID. The procedure is simple to perform, it takes the same amount of time, and is no riskier than CID. That being said, it would be reasonable to say that learning a new technique based on such weak evidence seems somewhat troublesome and over complicated. CID is still a reasonable course of action…for now. I hope further studies are conducted to evaluate resolution of the abscess but also the difference in pain and cosmetic outcome with the different techniques.
Alexander “Zander” Prewitt is an EM Resident, Class of 2020