Enterocutaneous fistula

This topic received some attention from the college in Question 15 from the first paper of 2021, and in Question 22 from the first paper of 2018. The examiners produced a haemodynamically unstable patient who "appears to have nasogastric feed emanating from his dehisced laparotomy wound". Basically, the management of this "disgusting malady", described by Guillaume Dupuytren as an "accidental anus", consists of trying to decreasing its output and waiting until the inevitable septic abdominal disaster settles down enough to do something surgically definitive. While waiting, one deals with abdominal collections, poor nutrition, eroding skin, angry relatives, and all manner of fluid/electrolyte bewilderment.

There is a vast amount of literature available on this complex topic, as there are as many publisheable opinions as there are eminent surgeons. The time poor exam candidate will want just one paper to read. Cohen et al (2016) seems to be the ideal option, as it is called "Complex surgical abdomen: what the nonsurgeon intensivist needs to know". Lundy & Fischer (2010) also do a good historical perspective, for those with infinite time on their hands. Fischer also teamed up with Evenson (2016) to write a contemporary account of these conditions which is more surgically minded.

Management of an enterocutaneous fistula takes several discrete stages, which remain essentially unchanged since Chapman et al described them in 1964Evenson (2016) proposes a somewhat modernised version of these, which is used for the discussion section in the answer of Question 22 from the first paper of 2018. These phases, with some modification, are:

  • Stabilisation where you resuscitate the septic shock
    • Fluid resuscitation should be conservative to prevent intestinal oedema
    • Vasopressor support should be careful to prevent intestinal ischaemia
  • Diagnosis where you confirm the location and extent of the fistula, by:
    • Methylene blue dyed feeds
    • Contrast "fistulogram" using water-soluble contrast and plain films or fluoroscopy (contrsat both rectal and oral)
    • Contrast CT, which has largely surpassed the fistulogram
  • Evaluation for spontaneous resolution where you decide whether this needs to go to theatre urgently, or whether you can sit on it for a month. Apart from imaging and historicla features, this includes
    • Nutritional assessment, incl. prealbumin, transferrin and clinical assessment
    • Attempted early surgery if you think you can't wait for it to get better on its own (and if the patient is likely to survive)
  • Trial of non-operative management if the impression is that surgery can be delayed
  • Management of infectious complications during conservative management:
    • Intra-abdominal faecal contamination initially - managed with broad-spectrum antibiotics (eg. peperacillin/tazobactam) as well as an antifungal if there are risk factors (malignancy, re-laparotomy, long term antibiotics)
    • Abdominal collections after a prolonged course of systemic antibiotics
  • Management of non-infectious complications during conservative management: 
    • Fluid losses particularly of high-output fistulas: strategies to decrease output include
      • avoidance of vac dressings
      • TPN
      • Octreotide, l​​​​​operamide, opiates
      • Diverting proximal stoma
      • Relief of distal bowel obstruction
    • Malnutrition: TPN and vitamin supplementation
    • Gut atrophy: low volume enteral feeds, or enteral nutrition given via the fistula
    • Electrolyte disturbance resulting from high stoma output (management of this depends mainly on vigilant replacement)
    • Skin excoriation due to spills (management of this requires a dedicated stoma service, or plastic surgical input to deal with tissue loss)
  • Definitive surgical management; or, when it becomes clear that this is not an option,
  • Palliative care

Definition and classification

A fistula by definition is a connection between two epithelial-lined organs. There are several possible ways to classify an enterocutaneous fistula:

  • Anatomically, i.e. which is the perforated organ (enterocutaneous, duodenocutaneous, etc) 
  • Simple vs. complex, where the simple fistulas have only one tract
  • Short vs. long, where the short fistulas have a tract length of less than 2cm
  • Volume of output: where high output (>500ml/day) is associated with poorer outcomes
  • "Control", where an uncontrolled fistula has evidence if sepsis, intra-abdominal infection or skin breakdown.

Obviously, not all of these features will be revealed by clinical examination, and some investigations will be required.

Diagnosis and investigations

Sensibly, Cohen et al (2016) report that 

"...visualizing drainage of intestinal contents from an operative incision or drain site clearly establishes a diagnosis of an ECF"

Which makes sense. However, in many situations the material coming out of a dehisced infected wound is, shall we say, equivocal. In other words, the ICU team and the surgical team may disagree substantially as to whether the wound is issuing feeds, faeces, bile, pus, liquified pancreatic tissue or whatever other delighful product. Clearly some objective methods are called for. Suspicion of a complicated fistula arise when the patient develops features of systemic sepsis, as in the scenario presented by Question 22 from the first paper of 2018. Then, imaging may be performed which can confirm and elaborate on this suspicion. Suggested imaging modalities are:

  • Methylene blue dye test to determine that the material coming out of the would is in fact nasogastric feeds; you mix the dye into the feeds and observe blue wound discharge. 
  • Fistulogram, a series of plain X-rays taken after administration or oral, NG or rectal contrast (this can define the tractreasonably well). Almost nobody does this any more, except when for whatever reason doing it fluoroscopically
  • CT with NG/oral/rectal contrast has now surpassed fistulography as the investigation of choice. Adjacent collections and all other sorts of circus in the abdomen will also be revealed by this. 

Additionally, some further data will be required to help make a decision regarding earlier surgery. 

Evaluation for spontaneous resolution

Is this going to get better by itself ? Factors associated with improved rates of spontaneous closure and mortality are listed in this table from Evenson & Fischer (2016):

factors which influence the outcome of enterocutaneous fistula

Thus, to know whether this thing will self-resolve, you need to:

  • Determine the anatomical location of the fistula and the state of the bowel with a CT
  • Agree on the aetiology (for instance, force the surgeons to admit that there were enterotomies during adhesiolysis)
  • Determine whether the patient is septic (i.e. assess blood cultures, haemodynamic stability and inflammatory markers)
  • Measure the output for some days
  • Assess the nutritional status of the patient, including albumin, prealbumin and transferrin

What happens when they operate

Let's say you decided that self-resolution is clearly not going to happen. What will happen if you take this patient back to theatre? Probably, also terrible things, but it will be unlikely that these things will be any more terrible than what awaits the patient if you treat them conservatively when it's clear that there is no hope for successful conservative management. Hutchins et al (2004) reported a  43% mortality after relaparotomy for abdominal sepsis, the same numbers as Edmunds et al (1960). This, according to the authors, was "unlikely to be greater than with nonoperative treatment of intraabdominal sepsis". However, things get worse the older you get. The same prospective observational study demonstrated a 78% mortality after relaparotomy in patients older than 75 years. 

Latifi et al (2012) offer an insight into the management of enterocutaneous fistula in a "hostile abdomen", i.e. when you haven't waited for everything to completely settle down and the inflamed tissues are still friable and suffering from impaired healing. The authors refer to it as a "disastroma", and "stoma city". What "early" means in relation to early surgery is not completely defined. The best information regarding this is probably from Fazio et al (1983), who demonstrated a favourable mortality window within the first 10 days after the first operation:

  • Within 10 days: 13% mortality
  • Between 11 and 42 days: 21% mortality
  • After 42 days:   11% mortality 

Useful strategies for early surgery include:

  • Enter through "nonviolated" areas, i.e don't go back in through the same scar/wound
  • Proximal stoma for diversion (i.e. don't put another anastomosis in there)
  • Resection of diseased bowel
  • Washout, including antibiotic lavage
  • Vacuum-assisted closure dressings  are controversial because they may increase the output of the fistula by sucking more material 
  • Exclusion of the fistula by taking it out of the path of enteric contents.

An excellent article by Edmunds et al (1960) illustrates some of the approaches for these exclusion operations:

exclusion operations for enterocutaneous fistula

After early surgery, or while waiting for the fire to die in the hostile abdomen, one needs to spend their time carefully managing the infectious and noninfectious complications of the fistula.

Management of infectious complications of enterocutaneous fistula 

Sepsis from intra-abdominal collections: the fistula emptying foul-looking material out of the dehisced wound is actually the good fistula. That one is decompressing into the ICU environment, safely out of the patient. The bad fistula is the quiet one, which you can't see or smell. That one is emptying into the abdominal cavity and causing sepsis by way of faecal contamination

Abdominal sepsis in the wake of this sort of enterocalypse is difficult to manage with antibiotics. All the surgical literature hides behind words like "broad spectrum" and "empiric" without suggesting specific agents or offering rationale for specific areas of  coverage. From an infectious disease point of view, the IDSA guidelines (Solomkin et al, 2003) are as follows: 

  • Extend the spectrum to cover for resistant organisms, because hospital-acquired abdominal sepsis is usually a problem created by by less susceptible organisms (Montravers et al, 1996). "Resistant pathogens were isolated from 70% ofthe patients who underwent reoperations for postoperative peritonitis." There is apparently no need to cover enterococci by default (as it seems to make no difference). No specific drug combinations can be recommended because "local nosocomial resistance paterns should guide therapy"
  • An anti-fungal drug (initially, fluconazole) is indicated after re-laparotomy, but probably not initially. The IDSA recommends you cover for fungi in the following circumstances:
    • malignancy
    • immunosuppression
    • re-laparotomy
    • solid organ transplant
    • recurrent post-operative infections

Intra-abdominal collections  develop because the antibiotics are effective. organisms become walled off from the immune system, trapped in lakes of nutrient fluid inside the abdomen. Guidelines for management of intra-abdominal infections (eg. Sartelli et al, 2017) emphasise source control but make a statement that some sources "can be managed without definitive source control if responding satisfactorily to antimicrobial therapy". 

What grows there? de Ruiter et al (2009)  surveyed the nightmarish zoo of organisms which grows in the abdominal fluid of complicated surgical infections, and found that:

  • Initially, it's mainly gram-negatives and anaerobes.  Half of the patients have Gram-negatives, of which half are E.coli, and in about 1/3rd of patients there are several Gram-negative species cultured. Colorectal perforations favour this mix of bugs; they are more rare in upper gastrointestinal fistulae.
  • Later, it is mainly Gram-positive organisms. "Over time, the prevalence of [Gram-negative aerobes] in abdominal fluid decreased from 117 patients (52.9%) in the first culture to one patient (6.7%) in week 4 (efficacy 87%). The prevalence of Gram-positive bacteria increased from 42.5 to 86.7% in a 4-week period.)
  • Upper GI fistulae are much more yeasty. The incidence of Candida was 41.0% in gastroduodenal perforations and 11.8% in colorectal perforations.
  • Lower GI perforations have more Gram-positive organisms,  presumable Clostridium and Enterococcus species.

Management of non-infectious complications

Basically all of these are due to the fact that bowel contents is not remaining in the bowel. By escaping before being turned into decent normal stool, fistula output carries nutrients electrolytes and fluids out with it, robbing the patient of valuable metabolic material. By spilling out onto the skin, the fistula output reminds us that it is in fact a part of the digestive system, by digesting its way into the skin and wound. These are te 

  • Fluid losses particularly of high-output fistulas: strategies to decrease output include
    • avoidance of vac dressings
    • TPN
    • Octreotide
    • Diverting proximal stoma
    • Relief of distal bowel obstruction
  • Malnutrition: TPN and vitamin supplementation
  • Gut atrophy: low volume enteral feeds, or enteral nutrition given via the fistula
  • Electrolyte disturbance resulting from high stoma output (management of this depends mainly on vigilant replacement)
  • Skin excoriation due to spills (management of this requires a dedicated stoma service, or plastic surgical input to deal with tissue loss)

Resolution or definitive management

When do you finally operate, if you took the conservative approach and the aptient survived it? That abdomen is never going to look perfect. Timing of surgery is hard to determine. According to Galie et al (2006)

"...There is no definitive way to tell when this dense reaction has subsided, but there are indications on careful clinical examination. The abdomen should be soft and nontender and the prior scar should be pliable. The abdominal wall should have healed as much as possible around the fistula and be free of inflammation..."

This area is unlikely to interest the intensivist, who (to generalise unfairly) is a creature of short-term interests, focused on the now. As such, events which will take place twelve months after your week are unlikely to give you food for thought, except as a topic for conversation during the family conference. More will not be spent on this matter in this already excessive summary chapter, except to point towards some high-quality revies (eg. Gribovskaja-Rupp et al, 2016)


Cohen, Wess A., et al. "The complex surgical abdomen: What the nonsurgeon intensivist needs to know." Journal of intensive care medicine 31.4 (2016): 237-242.

Edmunds Jr, L. Henry, G. M. Williams, and Claude E. Welch. "External fistulas arising from the gastro-intestinal tract." Annals of surgery 152.3 (1960): 445.

Lundy, Jonathan B., and Josef E. Fischer. "Historical perspectives in the care of patients with enterocutaneous fistula." Clinics in colon and rectal surgery 23.3 (2010): 133.

Chapman, Richard, Robert Foran, and J. Englebert Dunphy. "Management of intestinal fistulas." The American Journal of Surgery 108.2 (1964): 157-164.

Evenson, Amy R., and Josef E. Fischer. "Current management of enterocutaneous fistula.Journal of gastrointestinal surgery 10.3 (2006): 455-464.

Gribovskaja-Rupp, Irena, and Genevieve B. Melton. "Enterocutaneous fistula: proven strategies and updates." Clinics in colon and rectal surgery 29.2 (2016): 130.

Hutchins, Robert R., et al. "Relaparotomy for suspected intraperitoneal sepsis after abdominal surgery." World journal of surgery 28.2 (2004): 137-141.

Green, Gemma, et al. "Emergency laparotomy in octogenarians: A 5-year study of morbidity and mortality." World journal of gastrointestinal surgery 5.7 (2013): 216.

Solomkin, Joseph S., et al. "Guidelines for the selection of anti-infective agents for complicated intra-abdominal infections." Clinical Infectious Diseases 37.8 (2003): 997-1005.

Montravers, P., et al. "Emergence of antibiotic-resistant bacteria in cases of peritonitis after intraabdominal surgery affects the efficacy of empirical antimicrobial therapy." Clinical Infectious Diseases 23.3 (1996): 486-494.

Latifi, R., et al. "Enterocutaneous fistulas and a hostile abdomen: reoperative surgical approaches." World journal of surgery 36.3 (2012): 516-523.

Galie, Kathryn L., and Charles B. Whitlow. "Postoperative enterocutaneous fistula: when to reoperate and how to succeed." Clinics in colon and rectal surgery 19.4 (2006): 237.

Sartelli, Massimo, et al. "Management of intra-abdominal infections: recommendations by the WSES 2016 consensus conference." World journal of emergency surgery 12.1 (2017): 22.

De Ruiter, J., et al. "The epidemiology of intra-abdominal flora in critically ill patients with secondary and tertiary abdominal sepsis." Infection 37.6 (2009): 522.

Fazio, Victor W., Theodore Coutsoftides, and Ezra Steiger. "Factors influencing the outcome of treatment of small bowel cutaneous fistula." World journal of surgery 7.4 (1983): 481-488.

Dupuytren, G. "Memoir on a new method of treating accidental anus." Mem Acad Roy Med 259 (1828): 1828.