See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315719634

Treatment of Enterocutaneous Fistulas, Then and Now

Article  in  Nutrition in Clinical Practice · March 2017

DOI: 10.1177/0884533617701402

CITATIONS READS

22 1,121

9 authors, including:

Luis Alfonso Ortiz Reyes Bin Zhang

Massachusetts General Hospital Massachusetts General Hospital
23 PUBLICATIONS   87 CITATIONS    5 PUBLICATIONS   22 CITATIONS   

SEE PROFILE SEE PROFILE

Maureen McCarthy Haytham M A Kaafarani

Massachusetts General Hospital Massachusetts General Hospital
11 PUBLICATIONS   24 CITATIONS    352 PUBLICATIONS   3,838 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The Journal Club of the Student & Trainee Section of The American Society for Parenteral and Enteral Nutrition (ASPEN) View project

COVIDSurg Collaborative View project

All content following this page was uploaded by Bin Zhang on 02 November 2018.

The user has requested enhancement of the downloaded file.

701402
research-article2017
NCPXXX10.1177/0884533617701402Nutrition in Clinical PracticeOrtiz et al

Clinical Research
Nutrition in Clinical Practice
Volume XX Number X
Treatment of Enterocutaneous Fistulas, Then and Now Month 201X 1­–8
© 2017 American Society
for Parenteral and Enteral Nutrition
DOI: 10.1177/0884533617701402
https://doi.org/10.1177/0884533617701402
journals.sagepub.com/home/ncp
Luis Alfonso Ortiz, MSc1; Bin Zhang, PhD2;
Maureen Walsh McCarthy, MSN, CWON, AGNP-C3;
Haytham M. A. Kaafarani, MD, MPH1; Peter Fagenholz, MD1;
David R. King, MD1; Marc De Moya, MD1; George Velmahos, MD, PhD1;
and Daniel Dante Yeh, MD1

Abstract
Background: An enterocutaneous fistula (ECF) is an aberrant connection between the gastrointestinal tract and the skin or atmosphere
(enteroatmospheric fistula [EAF]). Multimodal treatment includes surgical procedures, nutrition support, and wound care. We evaluated
our practice and compared our outcomes with previous results published from our institution. Materials and Methods: We performed
a retrospective analysis of hospitalized ECF/EAF patients admitted between January 2011 and November 2015. Patients with internal
fistulas; active inflammatory bowel disease; malignancy; radiation treatment; end-stage renal, hepatic, or cardiac disease; and active
alcoholism were excluded. Data collected included demographics, fistula characteristics, nutrition therapy, treatment, operative success,
and hospital mortality. Parametric and nonparametric tests for independent and paired groups were performed. Results: Thirty-one
patients were included in the analysis. The median (interquartile range) age was 60 (53–76) years, and 81% were female. Parenteral
nutrition was initially prescribed in 80% of patients, but 61% received enteral nutrition (EN) at some point during their hospitalization.
Two patients were fed by fistuloclysis. Eighty percent of the patients underwent surgical repair a median of 12 months after diagnosis with
92% operative success. Surgical repair had a modest correlation with home discharge (ρ = 0.517, P = .003). A large proportion of patients
(77%) were discharged home. The in-hospital mortality at our institution decreased from 44% in 1960 to 21% in 1970 to 3% in the current
study. Conclusions: Modern treatment of ECF/EAF, including EN and advanced local wound care, is associated with improvements in
clinical outcomes such as hospital mortality. (Nutr Clin Pract. XXXX;xx:xx-xx)

Keywords
enterocutaneous fistula; fistula; enteroatmospheric fistulas; fistuloclysis; cutaneous fistula; intestinal fistula; enteral nutrition; parenteral
nutrition

Background enteral nutrition (EN) intake proximal to the fistula in an

ECF Definition, Etiology, Incidence, and
Morbidity
Enterocutaneous fistula (ECF) is defined as an abnormal, unin-
tentional connection between the gastrointestinal tract (GI) and
the skin or GI and the atmosphere (enteroatmospheric fistula
[EAF]) in the open abdomen.1 Most occur as a complication
following abdominal operation, although 15%–25% arise
spontaneously, typically in the setting of inflammatory bowel
disease (IBD), malignancy, radiation, or diverticulitis.2–4 The
development of an ECF is associated with many complications
such as prolonged hospital length of stay (LOS), increased
costs, electrolyte imbalance, and sepsis. Reported mortality
rates range from 6%–33%, with the most common causes of
death being attributable to malnutrition and sepsis.5–8 In survi-
vors, ECF/EAF has a strong negative impact on overall health-
related quality of life.8
Traditionally, standard treatment has involved correction of
fluid and electrolyte abnormalities, control of sepsis, and
“bowel rest.” Patients were commonly kept without any oral or
attempt to minimize fistula output. Nutrition therapy was
exclusively via the parenteral route, and calories were typically
prescribed in doses far exceeding estimated requirements, a
practice termed hyperalimentation. These practices may have
contributed to hyperglycemia and catheter-related infections,
From 1Massachusetts General Hospital, Department of Surgery, Division
of Trauma, Emergency Surgery, and Surgical Critical Care, Boston,
Massachusetts, USA; 2Massachusetts General Hospital, Department of
Pharmacy, Boston, Massachusetts, USA; and 3Massachusetts General
Hospital, Department of General Surgery, Boston, Massachusetts, USA.
Financial disclosure: D.D.Y. has received an educational grant from
Nestlé for an unrelated research study. No other financial disclosures are
declared.
Conflicts of interest: None declared.
Corresponding Author:
Daniel Dante Yeh, MD, Massachusetts General Hospital, Department of
Surgery, Division of Trauma, Emergency Surgery, and Surgical Critical
Care, 165 Cambridge St, #810, Boston, MA 02114, USA.
Email: daniel.dante.yeh@gmail.com
2 Nutrition in Clinical Practice XX(X)
which may partially account for the morbidity and mortality
seen in historical series.9 In the modern era, more moderate
parenteral nutrition (PN) caloric prescription is the standard,
and more effort is made to use the GI both proximal and distal
to the fistula. Available pharmacological agents such as loper-
amide, diphenoxylate/atrophine, octreotide, proton-pump
inhibitors, and cholestyramine can help decrease intrinsic
intestinal secretions as well as slow motility. Treatment with
hormonal mediators and peptide growth factors has also been
described with varying degrees of success.6,8 Surgical interven-
tions range from simple resection of the fistula with primary
reanastomosis to more complex surgical techniques involving
tissue transfer and abdominal wall reconstruction (with or
without biologic mesh) combined with negative-pressure
wound therapy (NPWT).3
Previously published series from our institution have
described the perioperative management of ECF/EAF, nutri-
tion therapy, wound care, medical and surgical treatment,
and outcomes, although the most recent publication was
from over 30 years ago.7,10,11 Recent publications from oth-
ers have reported drastically improved results.12,13 With the
evolving management of this challenging disease, we sought
to compare 21st-century nutrition therapy with historical
practices. We hypothesized that clinically important out-
comes such as operative success and survival were improved
in the modern era.
Methods
This study was approved by the local institutional research
board. We performed a retrospective chart review of adult
(age ≥18 years) patients with ECF/EAF treated at our institu-
tion from 2011 to 2015. Patients with internal fistulae (such
as colo-vesicular or gastro-gastric fistula) were excluded
because of the difficulty in quantification and determination
of treatment efficacy. In addition, patients with known risk
factors for poor healing of ECF/EAF (active IBD; active
malignancy; active radiation treatment; end-stage renal,
hepatic, or cardiac disease; and active alcohol consumption)
were excluded to avoid confounding. Demographic data col-
lected included age, baseline weight, body mass index (BMI),
and Charlson Comorbidity Index (CCI).5 Fistula characteris-
tics recorded included etiology, type, location, high-volume
(≥500 cc/24 hours) vs low-volume (<500 cc/24 hours) output,
and time elapsed from fistula onset to surgical repair.
Nutrition information collected included the nutrition method
provided (EN, PN, mixed EN and PN, or fistuloclysis), dura-
tion of artificial nutrition therapy, and amount of calories and
protein prescribed. Clinical outcomes recorded included
operative success rate, hospital LOS, in-hospital mortality,
and discharge disposition (home, rehabilitation, or nursing
home). Operative success was defined as no recurrence of
ECF within 90 days of operation.
Clinical Management
In general, our approach to ECF/EAF conformed to the 3-stage
approach described by Schecter et al.1 During phase 1, recog-
nition and stabilization, the patient was initiated on artificial
nutrition therapy and provided fluid rehydration with correc-
tion of electrolyte imbalances. The registered dietitians (RDs)
and nutrition support unit performed nutrition assessments in
all patients, estimating the caloric (kcal) and protein (g) goals
based on weight, BMI, and ideal body weight (IBW), from
25–40 kcal and 1.5–2.5 g. Supplemental PN was initiated in
patients presenting with obvious malnutrition, when it was felt
that tube feedings were not being adequately absorbed, or
when patients could not maintain adequate macronutrient
intake through per os (PO) or EN alone. For high-output fistu-
las (HOFs), PN was chosen as the first-line nutrition therapy to
ensure correction of hypovolemia. Adjustments to calorie and
protein prescription during the course of hospitalization were
made by the RD on a case-by-case basis using multiple sources
of evidence, including weight loss/gain, increase/decrease in
strength, calorie counts, indirect calorimetry, fistula output,
and metabolic biomarkers. A specialized wound ostomy nurse
was routinely consulted for assistance with control of fistula
effluent using ostomy bags, soft catheters, or NPWT. For EAF,
attempts were made to isolate the fistula from the surrounding
tissue using NPWT to “mature” the fistula into an ostomy.
Routine pharmacologic therapy included high-dose acid sup-
pression with a proton-pump inhibitor and antimotility agents
(loperamide and diphenoxylate/atropine) titrated to target a fis-
tula daily output <1500 mL or limited by side effects.
Cholestyramine and octreotide were selectively prescribed for
HOFs. This first stage generally lasted <1 week. In phase 2,
anatomical definition and decision, the location and length of
the fistula were determined using computed tomography (CT)
with oral contrast, fistulagram (contrast injection of the fis-
tula), or small bowel follow-through fluoroscopic examina-
tions. During this stage, attempts were made to initiate oral or
nasoenteric nutrition for distal fistulas. For proximal fistulas,
the fistula was cannulated, if possible, with a soft silastic feed-
ing tube that was directed distally. If successful, this tube was
secured to the skin and EN was provided into the distal GI
tract. When feasible, fistulocylsis was performed: the proximal
fistula output was collected and then reinfused through the dis-
tal feeding tube. The decision of whether or not to operate was
based on a multitude of factors such as whether the patient’s
fistula effluent was well controlled, frequency of septic epi-
sodes, fistula location, trajectory of fistula volume, patient’s
overall health status, and the patient’s preference. In general,
attempts were made to defer operation for at least 6 months.
For EAF, once the fistula had been isolated and the surround-
ing tissue developed into healthy, “beefy,” uncontaminated
granulation tissue, a split-thickness skin graft (STSG) was
applied to the surrounding tissue, essentially converting the
Ortiz et al 3

Figure 1.  Study flowchart of the selection process.

EAF into an ECF. Once the graft was fully reepithelialized, a Table 1.  Demographics, Fistula Characteristics, and Outcomes.
standard ostomy bag was used to control fistula output. In the
Characteristic Value
final phase, definitive operation, the patient was taken to the
operating room for exploratory laparotomy, excision of the fis- Female, No. (%) 25 (81)
tula, and primary reanastomosis. Abdominal wall reconstruc- Age, median (IQR), y 60 (53–76)
tion with component separation and biologic mesh was Baseline weight, median (IQR), kg 76 (64–91)
performed as necessary. Permanent prosthetic mesh was not Final weight, median (IQR), kg 73 (62–84)a
used. BMI, median (IQR), kg/m2 26.2 (22–30.2)
CCI, mean ± SD 2.3 ± 2.4
Type, No. (%)
Statistics  ECF 26 (84)
Descriptive statistics were calculated for demographic and  EAF 5 (16)
clinical characteristics. Continuous variables are presented as Location, No. (%)
mean (standard deviation) or median (interquartile range  Ileum 13 (42)
 Jejunum 12 (39)
[IQR]) for nonnormal distributions according to the
 Colon 6 (19)
Kolmogorov-Smirnov test. Student t test for independent and
Output, No. (%)
paired samples was calculated. Friedman test was used for
  High (≥500 cc/24 hours) 22 (71)
repeated measures, Spearman ρ was calculated and tested for
  Low (<500 cc/24 hours) 9 (29)
significance. A value of P < .05 was considered significant.
Treatment
SPSS version 20 (SPSS, Inc, an IBM Company, Chicago, IL)   Surgical, No. (%) 25 (81)
was used for statistical calculations.   Operative success 23/25 (92)
  Fistula onset to surgical repair, 12 (3–22)
Results median (IQR), mo
Hospital LOS, median (IQR), d 15 (10–38)
Patient Characteristics In-hospital mortality, No. (%) 1 (3)
Discharge, No. (%)
Between January 2011 and November 2015, a total of 69
 Home 24 (77)
patients with ECF/EAF were identified, and after removing
 Rehab 5 (16)
patients with 1 or more exclusion criteria, 31 patients were
  SNF/another hospital 1 (3)
included in the final cohort (Figure 1). Demographics, fistula  Died 1 (3)
characteristics, treatment, and clinical outcomes are summa-
rized in Table 1. Median (IQR) follow-up time from hospital BMI, body mass index; CCI, Charlson Comorbidity Index; EAF,
discharge until last clinical encounter for the entire cohort was enteroatmospheric fistula; ECF, enterocutaneous fistula; IQR,
interquartile range; LOS, length of stay; SNF, skilled nursing facility.
435 (174–854) days. Mortality after hospital discharge was a
P < .05 compared with basal weight.
16% (n = 5) and occurred a median (IQR) of 458 (98–747)
days after hospital discharge.
2). The median (IQR) time from fistula onset to operative repair
was 12 (3–22) months. Thirty of the 31 patients (97%) were
Patient Outcome
discharged alive, and the majority (77%) were discharged
Twenty-five patients (81%) underwent surgical repair, with a home. We found a moderate correlation for surgical repair and
92% operative success rate for those undergoing surgery (Figure home discharge, as demonstrated by the Spearman’s test (p =
4 Nutrition in Clinical Practice XX(X)

Figure 2.  Treatment and outcomes. HOF, high-output fistula; LOF, low-output fistula.

.517, P = .003). In the EAF group, 1 patient died of multisystem Table 2.  Characteristics of the Nutrition Support.
organ failure secondary to complications related to the EAF.
Characteristic Value
This patient had a prolonged intensive care unit (ICU) course
(nearly 3 months) and required multiple abdominal wall Initial nutrition support
debridements, vacuum assisted closure (VAC) changes, and tra-   PN, No. (%) 25 (81)
cheostomy. Ultimately, she developed a deep intra-abdominal    Days on PN, median (IQR) 50 (9–294)
infection that resulted in pseudoanerysm of her mesenteric   EN, No. (%) 4 (13)
artery. This developed into a life-threatening hemorrhage, and   EN + PN, No. (%) 2 (6)
there was no surgical or endovascular option to control the Fistuloclysis, No. (%) 3 (10)
hemorrhage. Therefore, the family decided to withhold contin- Home PN, No. (%) 18 (58)
ued life-sustaining treatments and transition to comfort care. PN prescription, mean ± SD
  Calories prescribed, kcal 1669 ± 319
  kcal/kg prescribed 21.5 ± 7
Nutrition Support Therapy   Grams of protein prescribed, g 94.8 ± 19
  g/kg prescribed 1.6 ± 0.36
Table 2 describes the nutrition therapy. The enteral route (PO,
Days to reach the goal, median (IQR) 5 (3–8)
EN, fistuloclisys) was used in 61% of our patients at some
point during their hospital stay. The majority (n = 25) received EN, enteral nutrition; HOF, high-output fistula; IQR, interquartile range;
PN as the initial nutrition support, which was initiated a median PN, parenteral nutrition.
(IQR) of 1 (1–5) day after hospital admission. Over half were
successful in transitioning to either EN or PO intake (Figure 3). compared with historical reports. In both periods, the princi-
Patients received artificial nutrition for a median (IQR) of 50 pal cause for ECF/EAF was a surgical complication. HOF
(9–294) days. Fistuloclysis was attempted in 3 patients, predominated in both periods, although the exact proportion
although only 2 were successfully fed with fistuloclysis. The was not reported in the Aguirre et al.11 The prescription of PN
average kcal and grams of protein initially prescribed by PN as initial nutrition therapy is still preferred, although the tran-
were 1199 ± 430 kcal and 79 ± 23 g or 19.9 ± 6 kcal/kg/d and sition to oral or EN was possible in most patients. In 1974,
1.3 ± 0.3 g/kg/d, respectively. Over a median (IQR) of 5 (3–8) the combination of PN and elemental diet was described in 1
days, the amounts of kcal and grams of protein prescribed patient, who received 6000 kcal/d. The previous initial daily
increased up to 1669 ± 319 kcal and 94.8 ± 19g or 21.5 ± 7 prescription of kcal was 4000–5000 kcal/d and 56–156 g
kcal/kg and 1.6 ± 0.36 g/kg (Table 2). protein/d; our current nutrition practice is to initially pre-
scribe 1199 kcal/d and 79 g protein/d, a much more moderate
initial prescription that avoids the risk of refeeding syndrome
Historical Comparison
and accounts for concomitant nutrition intake via EN or PO.
Surgical closure, nutrition support, and mortality.  Table 3 Most important, the overall mortality decreased from 21% in
summarizes some of the key similarities and differences 1974 to 3% in 2015, and the operative success rate increased
between our current clinical outcomes and nutrition therapy from 70% to 92%.11
Ortiz et al 5
Figure 3.  Nutrition therapy. EN, enteral nutrition; PN, parenteral
nutrition; PO, per os.
Discussion
The treatment of fistulas has advanced significantly since
Susruta, the father of Indian medicine (450 bc), described the
management of an ECF by resection and closure using the sev-
ered heads of ants.14 The treatment of ECF/EAF is very chal-
lenging and requires a multidisciplinary approach for optimal
outcomes. Our inpatient team includes surgeons, gastroenter-
ologists, interventional radiologists, RDs, pharmacists, and
advanced wound ostomy nurses. With collaboration, patience,
and expertise, we were able to achieve a high operative success
rate with a low mortality rate. A previous report from our same
institution from 1974 reported an operative success rate of
70%, although no information is given on the time elapsed
from fistula appearance to operative repair.11 In our practice of
waiting a median of 12 months, we report an operative success
rate of 92%. Initial operative success is important because fail-
ure leading to recurrence has been identified as a primary
determinant of mortality.13
Our results are in line with what other recent series have
reported. In a similar review of modern outcomes (compared
with a publication from the same institution), Draus et al15
reported a 7% overall mortality rate and 89% operative success
rate. A more recent series by Visschers et al16 described 79
patients with ECF treated from 2006–2010. They attempted
operation at a median of 14 weeks and reported a laudable 96%
operative success rate with an overall mortality rate of 10%.
Despite these encouraging results, our practice has been to wait
at least 6 months before attempted fistula resection/repair, as
recommended by Hollington et al.17
The main challenge of treating EAF is in managing fistula
output. Typically, these fistulae arise in the setting of an open
abdomen or when a closed incision is completely opened. In
this environment, the fistula contents simply pool up and ade-
quate wound healing is impossible in such an uncontrolled set-
ting. Ostomy appliances are universally unsuccessful because
they are unable to adhere to the un-epithelialized surrounding
tissues. Through collaboration with our wound ostomy nurse,
we employed multiple approaches to control the fistula efflu-
ent, including the creative use of specialized NPWT sponges
with higher tensile strength, increased density, and decreased
adherence (VAC WhiteFoam Dressings; KCI, San Antonio,
TX); skin barrier rings (Eakin Cohesive; ConvaTec,
Greensboro, NC); multiple types of soft tubes; and ostomy
bags. Early in the patients’ course, they required frequent VAC
changes, sometimes daily or more often, to maintain the intri-
cate dressings. However, with time and vigilance in maintain-
ing fistula effluent diversion, the wound became shallower and
easier to manage. Combined with adequate nutrition, the
wound matured into healthy granulation tissue and became
flush with the skin. STSG was then applied to convert an EAF
into an ECF. This allowed hospital discharge (to home or reha-
bilitation) while the patient waited for the appropriate time for
surgical intervention. In our sample, 40% of EAFs were treated
with STSG, with 100% successful graft take.
EN Therapy for EAF and ECF
“If the gut works, use it or prepare to lose it.”18,19 In a signifi-
cant proportion of our patients (61%), the enteral route was
employed during their hospital stay, a stark contrast from the
traditional dogma recommending bowel rest.4,8,10,20,21
Utilization of the GI tract to nourish ECF patients is not a new
concept, and others have reported successful use of EN.3,22,23 In
a review of 186 patients treated between 1968 and 1977, Reber
et al24 reported that in the latter period (1972–1977), 29% of
the patients did not require PN. Intraluminal nutrition stimu-
lates enhanced mesenteric perfusion, a well-known phenome-
non termed postprandial splanchnic hyperemia.25 Since
adequate perfusion is a requisite for wound healing, this mech-
anism may help explain the improved operative success rates.
Animal studies confirm superior results when patients are fed
by EN compared with isocaloric PN; the anastomotic bursting
pressure in animal subjects after GI surgery is higher and col-
lagen content is higher when early proximal EN is provided.26–28
Meta-analyses of human randomized trials of early EN after GI
operations also show a trend toward decreased risk of anasto-
motic dehiscence compared with control groups.29,30 Even if
unable to supply adequate nutrition, EN is believed to confer
multiple nonnutrition benefits.31 Using the enteral route is
6 Nutrition in Clinical Practice XX(X)

Table 3.  Comparison Between 1970 and 2016 of the Treatment and Outcomes for Enterocutaneous Fistula at MGH.a

Variable MGH, 1970 MGH, 2015

Etiology of ECF/EAF Surgical complication as Surgical complication as
principal cause: 77% principal cause: 93%
Split-thickness skin graft in EAF Not reported 40%
Fistula output HOF: predominantly in HOF: 71%
most of the cases
Type of treatment Surgical: 71% Surgical: 80.6%
Medical: 28% Medical: 19.4%
Time between fistula onset and operation Not reported 12 (3–22) months
Operative success 70.4% 92%b
Fistuloclysis Not reported 10%
Nutrition support initially prescribed PN (100%) PN (80.6%)
PN kcal prescribed/24 hours 4000–5000 1199 ± 430
Grams protein prescribed/24 hours 56–156 79.6 ± 2
EN kcal prescribed/kg/24 hours Not reported 27.5 (22–37)
EN grams protein prescribed/kg/24 hours Not reported 1.6 (1–2)
PN home Not available Yes
Since 1978
PN dependency Not reported 38%
Duration of PN dependency (years) Not reported 3 (2–7)
Overall mortalityc 21% 3%

EAF, enteroatmospheric fistula; ECF, enterocutaneous fistula; EN, enteral nutrition; HOF; high-output fistula; MGH, Massachusetts General Hospital;
PN; parenteral nutrition.
a
Values are reported as mean ± standard deviation or median (interquartile range) unless otherwise indicated.
b
Patients with conservative treatment and patients without plan to repair the ECF/EA were excluded.
c
Not defined in 1970 manuscript, defined as in-hospital mortality in 2015.

believed to enhance the functional and structural integrity of
the GI tract, prevent bacterial adherence to the epithelial cell in
gut, stimulate the secretion of immunoglobulin A, and support
the mass of gut-associated lymphoid tissue (GALT).32,33
Fistuloclysis was one option we used to provide nutrition
therapy for those fistulas not expected to close spontane-
ously.4,34 Distal enteric reinfusion of the fistula output is a tech-
nique called fistuloclysis that was first described in patients
with small intestinal fistula within an open abdomen.35
Depending on the level of the fistula, the effluent may be rich
in salivary amylase, gastric pepsin, pancreatic enzymes, and
bile. Restoration of the normal enterohepatic circulation may
improve liver function and in some cases may successfully lib-
erate the patient from PN altogether.35,36 The use of fistulocly-
sis has been shown to decrease the volume of the fistula output;
improve symptoms such diarrhea vomiting, nausea, abdominal
pain, and abdominal distention; improve liver function; and
improve nutrition status.23,37,38 Admittedly, our use of fistu-
loclysis was low. Many patients were not eligible because of
the distal location of the fistula or inability to cannulate the
fistula to obtain distal enteral access. For those who were eli-
gible, we sometimes encountered significant barriers from the
nurses and the patients, and several refused to reinfuse enteric
contents after they had left the body.
PN. There is universal consensus that EN is the preferred
method of nutrition therapy over PN if both options are
available. While we were aggressive in promoting the use of
the GI tract, a significant proportion of our patients still
required some degree of parenteral support to prevent dehydra-
tion and to provide adequate calories/protein. Initially hailed as
a lifesaver and godsend for patients with no other nutrition
options, the adverse effects of PN soon became apparent—
namely, hyperglycemia, catheter-related bloodstream infec-
tion, central vein thrombosis, and PN-associated liver disease
(PNALD). These accumulated complications engendered some-
times fierce reactions against the use of PN, even in patients with
legitimate need, and some have even suggested that the acronym
TPN should stand for “total poisonous nutrition.”39 However,
recent advances in clinical care have mitigated at least some of
these risks. Improved glycemic control and central catheter care
bundles have greatly decreased infectious complications; cycling
of PN and decreased soy fats administration have decreased the
risk of PNALD. Recently, a large, pragmatic trial comparing
early EN vs early PN was performed. Over 2000 patients were
randomized, and there was no difference seen in number of
infectious complications, incidence of organ failure, ICU LOS,
hospital LOS, or mortality.40 Thus, PN in the 21st century is
now safer than ever before, although it is still the more expen-
sive option by far.
About half of our patients were discharged with home PN,
compared with none in 1974, as home PN did not became
available at our institution until 1978. Comparing the 2 eras,
there was a significant reduction in the proportion of patients
Ortiz et al 7
with PN as initial nutrition support in contrast with the year
1974 (PN; 100%) and the duration that the patient received it:
2311 days vs 50 days.
In the previous era, calories and protein were prescribed
without reference to anthropometric measures such as BMI and
IBW. With the discovery of increased metabolic demand in
stressed states, clinicians sought to match calorie intake to esti-
mated high caloric needs and sometimes above calculated needs,
a practice termed hyperalimentation. For example, in a descrip-
tion of practice, “Nutrition was considered to be ‘adequate’ if
3000 calories or more were supplied daily.”24 Many believed
that hyperalimentation resulted in higher rates of spontaneous
closure and decreased mortality. It is likely that this practice con-
tributed to some of the complications of PN—namely, hypergly-
cemia, infections, and PNALD. In the 21st century, the PN
prescription in our institution is based on the weight, BMI, and
IBW, starting initially with <50% of the calculated requirements
and ramping up to goal in an average time of 5 days, in the range
of approximately 1700 kcal/d and 80 g protein/d.
ECFs are highly variable in location, fistula volume, incit-
ing cause, and metabolic stress, and the hospital course is often
marked by multiple infectious complications and multiple-
organ failure; these may place additional demands requiring
adjustment of nutrition therapy. The prescription of calories
and protein should be individualized to each patient and tai-
lored accordingly as the clinical condition changes. For exam-
ple, a baseline malnourished septic patient with a high-output
duodenal fistula will require a higher calorie/protein prescrip-
tion than a patient with a stable, low-output colonic fistula after
elective surgery. The variability in clinical presentations and
courses likely explains the high degree of variability in pre-
scribing practices.
At our institution, ECF mortality has decreased steadily
from 44% in 1960 to 21% in 1974 and finally 3% in 2015.7,11 It
is no longer considered a life-threatening illness but rather a
chronic condition, and most of our patients were discharged
home.
Limitations
Several limitations of our study must be acknowledged. First,
our results are based on retrospective data collection, and the
study design places our results at risk of bias. For example, we
could not discern the decision making driving conservative or
surgery management. Although the exact reasons are not
explicitly stated in the chart, it is clear that some patients were
at high risk for anesthetic complications (eg, 91-year-old
woman with congestive heart failure secondary to aortic steno-
sis and kidney failure requiring renal replacement therapy),
high risk for surgical complications (eg, 56-year-old woman
with distant history of radiation enteritis on anticoagulation for
thromboembolic complications; 79-year-old woman with bac-
teremia and severe malnutrition despite PN), both anesthetic
and surgical complications (eg, 90 year-old woman on
prednisone for autoimmune hepatitis), poor expected outcome
(eg, 46-year-old woman with gastroparesis and peritoneal scle-
rosis), or successful nonoperative treatment (eg, 69-year-old
woman with colocutaneous fistula treated successfully with a
“fistula plug” by interventional radiology). It must be empha-
sized that the retrospective design precludes us from conclud-
ing that EN and NPWT caused improved clinical outcomes.
Additional confounding factors (such as glycemic control)
may also influence wound healing. Our descriptive study con-
trasts current practice patterns from previous practice at our
institution and describes temporal changes in outcomes. In
addition, we did not specifically collect data on the nutrition
makeup (eg, percentage of carbohydrates, protein, fats) that
our patients received. Because of the fluctuating and overlap-
ping routes of nutrition (PN, EN, and PO), we felt that any
attempt to quantify such information would be inaccurate and
of limited utility. Likewise, we did not collect information
about complications related to the route of nutrition (PN vs
EN), although there is recent evidence to suggest that the risk
of infectious complications is equivalent.40 Second, we
acknowledge that the PN kcal and grams protein presented in
Table 3 for the modern period are lower than what is currently
recommended (up to 2 g/protein/d). This is likely a reflection
that some of the patients were also taking concomitant PO/EN,
as well as the fact that some of these patients were also at risk
for refeeding syndrome. However, the retrospective study
design precludes confirmation. Third, the study was conducted
at a single urban, academic referral center, and this may limit
the generalization of the findings. In addition, our patients can
be generally considered to have healthy remaining bowel
because we excluded patients with IBD and radiation-damaged
intestines from our analysis. Our results, therefore, cannot nec-
essarily be extrapolated to other patient populations. Finally,
the relatively small sample size makes it difficult to perform
sophisticated statistical analysis. Despite these shortcomings,
we feel that our findings are meaningful and important. Our
study describes our nutrition and local wound care approach to
ECF/EAF and reports improved clinical outcomes compared
with previous eras. Additional studies should be performed to
confirm these associations.
Conclusion
ECFs and EAFs are challenging to manage and require com-
mitment and dedication from a multitude of healthcare provid-
ers for optimal recovery. Close collaboration, aggressive use of
the enteral route (including fistuloclysis, if feasible), and
appropriately timed surgical repair can result in a high opera-
tive success rate and a high rate of discharge home.
Statement of Authorship
L. A. Ortiz and D. D. Yeh equally contributed to the conception
and design of the research. All authors contributed to the acquisi-
tion, analysis, and interpretation of the data; drafted the
8 Nutrition in Clinical Practice XX(X)
manuscript; critically revised the manuscript; agree to be fully
accountable for ensuring the integrity and accuracy of the work;
and read and approved the final manuscript.
References
1. Schecter WP, Hirshberg A, Chang DS, et al. Enteric fistulas: principles of
management. J Am Coll Surg. 2009;209(4):484-491.
2. Datta V, Windsor AC. Surgical management of enterocutaneous fistula.
Br J Hosp Med (Lond). 2007;68(1):28-31.
3. Evenson AR, Fischer JE. Current management of enterocutaneous fistula.
J Gastrointest Surg. 2006;10(3):455-464.
4. Majercik S, Kinikini M, White T. Enteroatmospheric fistula: from soup to
nuts. Nutr Clin Pract. 2012;27(4):507-512.
5. Owen RM, Love TP, Perez SD, et al. Definitive surgical treatment of
enterocutaneous fistula: outcomes of a 23-year experience. JAMA Surg.
2013;148(2):118-126.
6. Williams LJ, Zolfaghari S, Boushey RP. Complications of enterocutaneous fis-
tulas and their management. Clin Colon Rectal Surg. 2010;23(3):209-220.
7. Edmunds LH Jr, Williams GM, Welch CE. External fistulas arising from
the gastro-intestinal tract. Ann Surg. 1960;152:445-471.
8. Polk TM, Schwab CW. Metabolic and nutritional support of the
enterocutaneous fistula patient: a three-phase approach. World J Surg.
2012;36(3):524-533.
9. Wang GF, Ren JA, Jiang J, Fan CG, Wang XB, Li JS. Catheter-related
infection in gastrointestinal fistula patients. World J Gastroenterol.
2004;10(9):1345-1348.
10. Soeters PB, Ebeid AM, Fischer JE. Review of 404 patients with gastrointes-
tinal fistulas. Impact of parenteral nutrition. Ann Surg. 1979;190(2):189-202.
11. Aguirre A, Fischer JE, Welch CE. The role of surgery and hyperali-
mentation in therapy of gastrointestinal-cutaneous fistulae. Ann Surg.
1974;180(4):393-401.
12. Lundy JB, Fischer JE. Historical perspectives in the care of patients with
enterocutaneous fistula. Clin Colon Rectal Surg. 2010;23(3):133-141.
13. Brenner M, Clayton JL, Tillou A, Hiatt JR, Cryer HG. Risk fac-
tors for recurrence after repair of enterocutaneous fistula. Arch Surg.
2009;144(6):500-505.
14. Bhishagratna KKL. An English translation of The Sushruta Samhita:
based on original Sanskrit text. Calcutta, India: Wilkens Press; 1907.
http://www.archive.org/stream/englishtranslati01susruoft#page/n47/
mode/2up/search/ant. Accessed February 1, 2011.
15. Draus JM Jr, Huss SA, Harty NJ, Cheadle WG, Larson GM. Enterocutaneous
fistula: are treatments improving? Surgery. 2006;140(4):570-578.
16. Visschers RG, an Gemert WG, Winkens B, Soeters PB, Olde Damink SW.
Guided treatment improves outcome of patients with enterocutaneous fis-
tulas. World J Surg. 2012;36(10):2341-2348.
17. Hollington P, Mawdsley J, Lim W, Gabe SM, Forbes A, Windsor
AJ. An 11-year experience of enterocutaneous fistula. Br J Surg.
2004;91(12):1646-1651.
18. Langkamp-Henken B. If the gut works, use it: but what if you can’t? Nutr
Clin Pract. 2003;18(6):449-450.
19. Baskin WN. Advances in enteral nutrition techniques. Am J Gastroenterol.
1992;87(11):1547-1553.
20. Rose D, Yarborough MF, Canizaro PC, Lowry SF. One hundred and four-
teen fistulas of the gastrointestinal tract treated with total parenteral nutri-
tion. Surg Gynecol Obstet. 1986;163(4):345-350.
View publication stats
21. Sitges-Serra A, Jaurrieta E, Sitges-Creus A. Management of postoperative
enterocutaneous fistulas: the roles of parenteral nutrition and surgery. Br J
Surg. 1982;69(3):147-150.
22. Makhdoom ZA, Komar MJ, Still CD. Nutrition and enterocutaneous fistu-
las. J Clin Gastroenterol. 2000;31(3):195-204.
23. Levy E, Frileux P, Cugnenc PH, et al. High-output external fistulae of the
small bowel: management with continuous enteral nutrition. Br J Surg.
1989;76(7):676-679.
24. Reber HA, Roberts C, Way LW, Dunphy JE. Management of external
gastrointestinal fistulas. Ann Surg. 1978;188(4):460-467.
25. Revelly JP, Tappy L, Berger MM, Gersbach P, Cayeux C, Chiolero R.
Early metabolic and splanchnic responses to enteral nutrition in postopera-
tive cardiac surgery patients with circulatory compromise. Intensive Care
Med. 2001;27(3):540-547.
26. Tadano S, Terashima H, Fukuzawa J, Matsuo R, Ikeda O, Ohkohchi N.
Early postoperative oral intake accelerates upper gastrointestinal anasto-
motic healing in the rat model. J Surg Res. 2011;169(2):202-208.
27. Kiyama T, Efron DT, Tantry U, Barbul A. Effect of nutritional route on
colonic anastomotic healing in the rat. J Gastrointest Surg. 1999;3(4):441-
446.
28. Fukuzawa J, Terashima H, Ohkohchi N. Early postoperative oral feeding
accelerates upper gastrointestinal anastomotic healing in the rat model.
World J Surg. 2007;31(6):1234-1239.
29. Andersen HK, Lewis SJ, Thomas S. Early enteral nutrition within 24h of
colorectal surgery versus later commencement of feeding for postopera-
tive complications. Cochrane Database Syst Rev. 2006;(4):CD004080.
30. Lewis SJ, Egger M, Sylvester PA, Thomas S. Early enteral feeding versus
“nil by mouth” after gastrointestinal surgery: systematic review and meta-
analysis of controlled trials. BMJ. 2001;323(7316):773-776.
31. McClave SA, Martindale RG, Rice TW, Heyland DK. Feeding the criti-
cally ill patient. Crit Care Med. 2014;42(12):2600-2610.
32. McClave SA, Heyland DK. The physiologic response and associated
clinical benefits from provision of early enteral nutrition. Nutr Clin Pract.
2009;24(3):305-315.
33. Genton L, Cani PD, Schrenzel J. Alterations of gut barrier and gut micro-
biota in food restriction, food deprivation and protein-energy wasting.
Clin Nutr. 2015;34(3):341-349.
34. Willcutts K, Mercer D, Ziegler J. Fistuloclysis: an interprofessional
approach to nourishing the fistula patient. J Wound Ostomy Continence
Nurs. 2015;42(5):549-553.
35. Teubner A, Morrison K, Ravishankar HR, et al. Fistuloclysis can success-
fully replace parenteral feeding in the nutritional support of patients with
enterocutaneous fistula. Br J Surg. 2004;91(5):625-631.
36. Picot D, Layec S, Dussaulx L, Trivin F, Thibault R. Chyme reinfusion
in patients with intestinal failure due to temporary double enterostomy: a
15-year prospective cohort in a referral centre [published online April 28,
2016]. Clin Nutr.
37. Wu Y, Ren J, Wang G, et al. Fistuloclysis improves liver function and
nutritional status in patients with high-output upper enteric fistula.
Gastroenterol Res Pract. 2014;2014:941514.
38. Ham M, Horton K, Kaunitz J. Fistuloclysis: case report and literature
review. Nutr Clin Pract. 2007;22(5):553-557.
39. Marik PE, Pinsky M. Death by parenteral nutrition. Intensive Care Med.
2003;29(6):867-869.
40. Harvey SE, Parrott F, Harrison DA, et al. Trial of the route of early nutri-
tional support in critically ill adults. N Engl J Med. 2014;371(18):1673-1684.