Management of patients with ECFs and abdominal wall defects remains complex and choices of repair may be limited due to the presence of local contamination from fistulas, which particularly restrict use of prosthetic mesh. Moreover, often these patients have concomitant malnutrition and may have other comorbidities. All these factors can increase perioperative morbidity if surgery is undertaken without optimal preoperative optimization and careful planning (see Chap. 26 on perioperative optimization). Optimization can only be achieved using a multidisciplinary approach and advanced surgical techniques. In addition, in our opinion, such a repair should only be attempted by those with clinical and surgical interest and expertise in complex ventral hernia repair. Often these patients have both fistulas and abdominal defects, and thus, even in the face of most optimal preoperative optimization, there is a question of whether a takedown of fistulas and complex abdominal wall defect reconstruction should be performed at the same time or separately. Furthermore, what surgical techniques and what mesh to use for repair of a large hernia defect in a contaminated field are other important questions that we do not have enough evidence to answer it. Subsequently, deciding on timing of surgery is not easy and needs to be individualized. Factors such as patient nutritional status, septic complications, comorbidities, and anatomy of the surgical problem play a major role in decision-making and should be analyzed carefully before definitive surgery.

The majority (75–85%) of ECFs are postoperative. Spontaneous ECFs are less common and typically occur in face of inflammatory bowel disease, malignancy, radiation, and diverticulitis. The etiology of a fistula has been shown to be predictive of the outcome. The worse outcome is seen in those with malignancy or trauma as the etiology [1]. Spontaneous closure rate of 20–37%, for patients with ECF has been achieved in recent large series [2]. The acronym “FRIENDS” is very useful to identify fistulas, which are unlikely to close spontaneously. Friends stands for: Presence of Foreign body, prior Radiation exposure, presence of inflammation (IBD) or Infection, Epithelization of tract, Neoplasm, Distal obstruction, and Sepsis/Steroids, all of which make it difficult for these fistulas to close spontaneously [3]. An ECF within an abdominal wall defect and not adjacent to viable skin the so called wide-mouth eneterocutaneous fistula is unlikely to close spontaneously. Spontaneous closure occurs most likely in patients with a closed abdomen, low output ECF, and an uncomplicated disease course. In patients with favorable factors, prolongation of the waiting period before surgery allows spontaneous closure rate to almost double from 16 to 29% [4].

With the popularization of damage control laparotomy (DCL), ECF and EAFs have become common problem in survivors, who require complex abdominal wall reconstruction [5]. Incidence of enterocutaneous fistula has been reported to be between 5 and 75%, and mortality can be as high as 40% in this subset of patients [6]. The loss of abdominal wall domain associated with such a large defect renders these patients unable to be fully functional. Although DCL has been shown to save lives, its overuse has consequences in the form of complex abdominal wall defects as well as a high risk of enteroatmospheric fistulas , and should be limited both for its use, and in particular, the length of time that the abdomen is left open [7]. Simply performing DCL should not commit the patient to open abdomen management. In setting of the open abdomen, most ECFs and EAFs develop early in the postoperative course (usually within 1 month). Small bowel fistulas are usually followed by large bowel and stomach fistulas. Likely etiologies in the open abdomen include anastomotic leak, mesh erosion into the bowel wall, exposed viscera, bowel injury during dressing changes, and splitting of the intestine from vigorous coughing or pulmonary toilet [8]. Significant predictors of ECF or EAF in patients with an open abdomen after DCL include large bowel resection, large-volume fluid resuscitation, and an increasing number of abdominal re-explorations [9]. These patients need to be stabilized hemodynamically and functionally with long-term nutritional support, skin graft coverage of open abdomen granulation tissue, and physical therapy before being considered for repair. Usually component separation or bridged mesh repair is required for such large abdominal wall defects.

In patients with Crohn’s disease , medical treatment itself may result in closure of ECF and abdominal wall repair can be attempted later. In general, however, the diseased segment of bowel resulting in the fistula requires surgical resection after a period of watchful waiting to allow intraabdominal inflammatory process to improve [10].

Radiation enteritis is a rare cause of ECFs, but is associated with poor prognosis. It has been advocated to strongly consider diverting proximal stomas, intestinal bypass, and resection with re-anastomosis using radiation spared bowel. If a fistula develops in the face of radiation enteritis, it is important to consider more conservative approaches. Patients are likely to have multiple strictures, malabsorption, and healing is likely to be poor. Most patients in this condition will end up needing long-term parenteral nutrition. Any abdominal surgery is associated with high risk of complications and the best outcomes are likely to be achieved by experienced surgeons operating on nutritionally optimized patients [11].

Timing of repair of enterocutaneous fistulas remains controversial and there is no Level 1 data available. An operative repair plan and timing needs to be individualized for every patient and every condition. If the index operation was performed more than 2 weeks prior, the most advantageous timing is to wait for 3 months to allow inflammation to settle down before attempting any major repair. Operating within 2 weeks or after 3 months has been shown to be associated with lower fistula recurrence rate compared to if surgery is performed between 2 and 12 weeks [17]. The only indication for surgery during this period would be complete SBO, bowel gangrene, peritonitis, or bleeding. Even under these circumstances, it should be limited to proximal defunctioning stoma formation and control of sepsis. This period should be used for improving nutritional status of patient, as it would have significant effect on morbidity/mortality associated with repair. Improved nutrition would also allow a significant percentage of fistulas to undergo spontaneous closure, simplifying subsequent abdominal wall defect repair surgery.

Once inflammation has attenuated, sepsis is controlled, nutritional status is optimized, and anatomy of the fistula is defined, operative repair should be attempted by an experienced surgeon. It has been suggested that these patients are managed better in specialized units, a suggestion with which we fully agree [18]. Specialized units with well-established protocols for stepwise staged management are much better equipped to handle care of these sick patients. Communication with patients and family is very important, as they need to understand each step with regard to goals and expectations.

We do not recommend waiting too long for repair beyond 3 months and surgery should be undertaken if conditions are favorable. Lynch et al. did argue for delaying surgery beyond 12 months to improve outcomes in patients with ECF [17]. However, Brenner et al. documented a five times increased risk of recurrence if operative repair was delayed beyond 36 weeks [19]. Owen et al. published similar results that showed that delaying surgery for longer than 1 year doubled the risk of postoperative refistulization [20]. Waiting too long also increases the risk of complications such as central line associated blood stream infections in patients on parenteral nutrition. Visschers et al. have recommended titrating the timing of restorative surgery to day-to-day patient characteristics after an initial waiting period of at least 6 weeks, but in their subsequent analysis, prolongation of the period of convalescence to a median of 101 days from 53 days was associated with an increased rate of spontaneous closure and a reduced recurrence rate after surgery [2, 4]. Rahbour et al. showed improved healing rates after surgery from 94.6 to 82% after increase in waiting time period to 12 months from 8 months previously [21]. Gupta et al. reported good results in their series when early interventions were carried out for ECF within 3 weeks of recognition. This may be a better strategy in resource poor settings where TPN may not be affordable for the long term, and delay in surgery can only worsen patient health status [22].