190

Operative Management of Anastomotic Leaks

after Colorectal Surgery
Nicole M. Saur, MD, FACS, FASCRS1 E. Carter Paulson, MD, MSCE, FACS, FASCRS1,2

1 Division of Colon and Rectal Surgery, Department of Surgery, Address for correspondence Nicole M. Saur, MD, FACS, FASCRS,
Perelman School of Medicine University of Pennsylvania, Division of Colon and Rectal Surgery, Department of Surgery,
Philadelphia, Pennsylvania Perelman School of Medicine University of Pennsylvania, 800 Walnut
2 Department of General Surgery, Corporal Michael J. Crescenze VA Street, 20th Floor, Philadelphia, PA 19107
Medical Center, Philadelphia, Pennsylvania (e-mail: [email protected]).

Clin Colon Rectal Surg 2019;32:190–195.

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Abstract Anastomotic leak is associated with increased morbidity and mortality after colorectal
surgery. Although surgical techniques have improved over time, anastomotic leak is
still a reality in colorectal surgery with rates ranging from as low as 1% for low-risk
anastomoses, such as enteroenteric or ileocolic, to 19% for high-risk coloanal anasto-
moses. There are many varied risk factors for anastomotic leak. However, many of the
risk factors have not been definitively proven in high-quality studies. Presumably, risk
factors are cumulative and every effort should be made to optimize modifiable risk
factors in the perioperative period. Treatment of anastomotic leak should start with the
Keywords determination of patient stability followed by resuscitation and diagnostic imaging or
► anastomosis operative exploration. Operative findings will dictate surgical approach with the goal of
► leak controlling sepsis and stabilizing the patient. If nonoperative treatment is undertaken,
► sepsis close patient monitoring is necessary to ensure control of sepsis and that intervention
► drainage is undertaken if the clinical picture changes. Early intervention at each stage is key to
► exploration decreasing the morbidity of anastomotic leak.

Rates and Impact of Anastomotic Leak slightly more than half of respondents agreed that “intra-
abdominal sepsis requiring a laparotomy,” or radiological
Although medical and surgical care of colorectal diseases collections treated with antibiotics or percutaneous drainage
have continued to evolve and improve, anastomotic leak constituted a leak.2 Bruce et al confirmed that there is no
rates remained relatively stable over time. A recent systema- standard anastomotic leak definition in the literature.3
tic review revealed an overall leak rate of 1 to 19% in color- Contributing to the heterogeneity is the varying post-
ectal surgery. Rates of anastomotic leak vary based on the operative time frame when leaks become apparent. Historic
type of anastomosis with enteroenteric having the lowest teaching has dictated that leaks typically present within
leak rate (1–2%) and colorectal/coloanal having the highest 1 week postoperatively. However, a large prospective review
(5–19%). Ileocolic (1–4%), colocolic (2–3%), ileorectal (3–7%), of 1,223 patients operated on by two surgeons over a 10-year
and ileoanal pouch (4–7%) anastomoses had intermediate period revealed a 2.7% leak rate and leaks were diagnosed at a
leak rates.1 mean of 12.7 days postoperatively. Four patients (12.1%)
There is no consensus definition of what constitutes a leak were diagnosed with a leak after 30 days. Forty-two per
and there is heterogeneity in surgeons’ own definitions both cent of leaks were diagnosed after the patient was dis-
from an academic and clinical perspective. Adams and Papa- charged home.4 These late leaks may not be diagnosed at
grigoriadis performed a survey of surgeons and definitions the primary institution, affecting database-driven analysis of
varied from “extravasation of contrast on enema” (94.2%) to leak rates, and may not be clearly defined as leaks and instead
“fecal material seen in drains or from the wound” (91.8%). Only treated as simple abscesses.

Issue Theme Complications and Copyright © 2019 by Thieme Medical DOI https://doi.org/
Dilemmas in Colorectal Surgery; Guest Publishers, Inc., 333 Seventh Avenue, 10.1055/s-0038-1677025.
Editor: Skandan Shanmugan, MD New York, NY 10001, USA. ISSN 1531-0043.
Tel: +1(212) 584-4662.
 Operative Management of Anastomotic Leaks after Colorectal Surgery Saur, Paulson 191

Complication rates and mortality rates are higher in showed an increased risk of anastomotic leak with grade 2
patients who have an anastomotic leak. Gessler et al evaluated anastomosis (odds ratio [OR]: 4.09, 95% CI: 1.21–13.63,
600 consecutive patients from 2010 to 2012 who underwent p ¼ 0.023). The authors concluded that revision of grades
colorectal anastomosis for colorectal cancer, diverticulitis, 2 and 3 anastomoses should be considered intraoperatively.8
inflammatory bowel disease, or benign polyps. They found a A retrospective review of 382 patients undergoing color-
10% leak rate and leaks were diagnosed at a mean of 8.8 days ectal resection evaluated the impact of the number of staple
(range: 2–42) postoperatively. Seventy-three per cent of fires required for rectal division on leak rates. In this study,
patients with leaks had an abdominal computed tomography one stapler load was used to divide the rectum in 58.4% of
(CT) scan and, of which, 25% were initially negative for leak. cases, two in 33.5% of cases, and three or more in 8.1% of
Seventy-six per cent of patients with leak required takedown cases. Male gender, history of neoadjuvant therapy, diagnosis
of their anastomosis. The overall complication rate (93.3 vs. of rectal cancer, laparoscopy, and operative duration more
28.5%, p < 0.001) and 30-day mortality rate (5 vs. 0.6%, than 200 minutes were the most prevalent reasons for
p ¼ 0.015) were higher in those patients who had a leak.5 In multiple stapler firings. The overall leak rate was 4.7%. The
a prospective study of 1,772 patients with cancer, the total only factor associated with increased risk of anastomotic leak

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
leak rate was 5.1%. In those patients diagnosed with a leak, the was three or more stapler loads used to divide the rectum.9
5-year overall survival rate was 44.3 versus 64% in patients Murray et al evaluated the association between laparoscopic
without a leak. In controlling for patient, surgical, and cancer- and open approaches on leak rate in 23,568 patients under-
specific factors, anastomotic leak had an independent negative going elective colorectal resections identified in the ACS-NSQIP
association with overall survival (hazard ratio [HR]: 1.6, con- (American College of Surgeons National Surgical Quality
fidence interval [CI]: 1.2–2.0) and cancer-specific survival (HR: Improvement Program) database. The overall leak rate was
1.8, CI: 1.2–2.6).6 3.4%; 2.8% in the laparoscopic group and 4.5% in the open group.
This significant difference persisted when controlling for
patient, disease, and procedure-related variables using pro-
Risk Factors for Anastomotic Leak
pensity-score matching (OR: 0.73, 95% CI: 0.58–0.91) and
Multitudes of risk factors for anastomotic leak have been multivariate analysis (OR: 0.69, 95% CI: 0.58–0.82).10 Two
described in the literature. In fact, a PubMed search using the randomized trials, however, did not show a difference in leak
terms “colorectal surgery,” “anastomotic leak,” and “risk rates between laparoscopic and open operations, although
factors” returns almost 500 articles. However, most of the surgeons were likely earlier in their laparoscopic learning curve
studies are low quality, many studies have shown conflicting in those series.11,12
results, and few factors have been shown to be definitively Not surprisingly, ischemia of the anastomosis has been
associated with anastomotic leak. Potential modifiable risk associated with anastomotic leak.13 Techniques such as fully
factors, however, should be minimized perioperatively as the mobilizing the proximal bowel (i.e., takedown of the splenic
effect of possible risk factors is likely cumulative. flexure) and performing appropriate vascular ligation (i.e.,
high ligation of the inferior mesenteric artery and vein) for a
Technical Risk Factors tension-free anastomosis are paramount. Dividing the
The location of the anastomosis, defined as intraperitoneal mesentery or mesocolon sharply at the time of bowel trans-
versus extraperitoneal, has been shown to be directly related ection has been described to verify adequate blood supply to
to the risk of leak. One of the largest prospective studies to the bowel prior to anastomosis. After mesenteric division,
examine this issue included almost 1,600 patients who confirming the presence of arterial blood supply can be aided
underwent 1,639 anastomoses. The risk of anastomotic by palpating a pulse or using a Doppler to confirm a signal.
leak following extraperitoneal anastomosis was 6.6% com- Newer techniques such as fluorescence angiography have
pared with 1.5% following intraperitoneal anastomoses. been advocated to provide visual confirmation of blood
More distal anastomoses had higher leak rates; ultralow or supply before and after performing an anastomosis.14
coloanal anastomoses had the highest leak rate (8%).7 The association between handsewn versus stapled ana-
The visual appearance of the anastomosis intraopera- stomoses and leak rates is varied in the literature and differs
tively can serve as a guide to identify anastomoses at high by site of anastomosis. A Cochrane review of seven trials
risk for leaking. Sujatha-Bhaskar et al evaluated 106 stapled showed that handsewn ileocolic anastomoses were asso-
colorectal anastomoses intraoperatively and graded them as ciated with a higher leak rate than stapled ileocolic anasto-
grade 1: circumferentially normal appearing mucosa, grade moses, for both benign and cancer diagnoses.15 For colorectal
2: ischemia or congestion of <30% of the colon or rectal anastomoses, however, stapled and handsewn anastomoses
mucosa, or grade 3: ischemia or congestion of >30% of the had similar leak rates, both clinically and radiologically.16
colon or rectal mucosa or ischemia/congestion of both sides Finally, the use of diverting ostomies has been evaluated
of the staple line. Any anastomoses (4/106) that were graded as a means of reducing the risk of anastomotic leak. Almost
3 were taken down and redone. Of the remaining cohort, 92 all studies have demonstrated that diverting ostomy reduces
of 106 anastomoses were graded as 1 and 10 of 106 were the clinical impact of a leak (need for reoperation, mortality,
graded as 2. The anastomotic leak rate for the cohort was etc.), but fewer have shown a clearly lower leak rate with a
12.2% with a leak rate of 9.4% for grade 1 anastomoses and diverting ostomy. In a 2014 prospective study of 2,364
40% for grade 2 anastomoses. Multivariate logistic regression patients undergoing 2,994 anastomoses, the use of a

Clinics in Colon and Rectal Surgery Vol. 32 No. 3/2019

192 Operative Management of Anastomotic Leaks after Colorectal Surgery Saur, Paulson

diverting stoma was not associated with decreased leak rate, with postoperative complications has been scrutinized. A
but was associated with decreased clinical severity of leak- systematic review and meta-analysis was published in 2016
age and decreased need for reoperative exploration.17 This and evaluated 6 randomized controlled trials (including 473
finding confirmed that of a previous prospective study.18 In a patients) and 11 observational studies (including more than
meta-analysis that included four randomized trials, how- 20,184 patients). NSAID use was associated with a significantly
ever, protective stoma was associated with both a lower leak increased risk of anastomotic leak in the observational studies
rate (9.6 vs. 22.8%) and a decreased need for reoperation (OR: 1.46, 95% CI: 1.14–1.86, I ¼ 54%). In the randomized
when a leak occurred (OR: 0.27, 95% CI: 0.17–0.59) compared controlled studies, however, NSAID use was not significantly
with those patients without stomas.19 associated with leaks (risk ratio: 1.96, 95% CI: 0.74–5.16,
I ¼ 0%). The pooled incidence of leak was 2.5% in patients
Patient Factors without NSAID usage, but 5.2% for NSAID users, which suggest a
A myriad of patient factors have been identified as risk difference clinically, although it was nonsignificant statisti-
factors for anastomotic leak, but few have been shown to cally.26 A NSAID of specific interest is ketorolac, which is often
be definitively or consistently associated with increased leak used intra- and postoperatively as part of enhanced recovery

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
rates. Several large, retrospective studies have identified pathways. One recent study (55% of included patients were
several modifiable and nonmodifiable patient factors possi- colorectal patients) showed that ketorolac use was associated
bly related to increased risk of anastomotic leak.20,21 with a higher rate of readmissions related to anastomotic
These include: complications (OR: 1.20, 95% CI: 1.06–1.36).27 However, an
earlier study of 731 patients with primary anastomoses, 48.6%
• Diabetes, perioperative hyperglycemia, and elevated
of whom received ketorolac perioperatively, showed no differ-
hemoglobin A1c
ence in leak rate with the addition of ketorolac (3.2 vs. 3.4%, OR:
• American Society of Anesthesiologists score 3
1.06, 95% CI: 0.43–2.62; p ¼ 0.886).28 Given the frequency of
• Young age
their use in enhanced recovery pathways in colorectal surgery,
• Smoking
additional well-powered randomized studies are necessary to
• Serum albumin <4, weight loss
determine if NSAID use truly increases anastomotic leak rate.
• Anemia
Overall, there are many technical and patient factors that
• Blood transfusion
likely relate, in a cumulative manner, to the risk of anasto-
• Neoadjuvant radiation
motic leak. First and foremost, in our opinion, surgeons
• Perianastomotic drain placement
should perform a tension-free anastomosis with no evidence
• Mechanical bowel preparation
of vascular compromise. The approach they choose, whether
• Increased operative time
laparoscopic or open, or handsewn or stapled should be an
• Emergency surgery.
approach that they feel most comfortable performing safely.
We explore a few other specific patient-related factors If there are concerns about the integrity of the anastomosis,
below. the anastomosis should be redone at the index operation. If
Male gender was found to be a risk factor for extraper- the anastomosis or the patient is deemed to have several risk
itoneal anastomotic leaks.20 This is presumably secondary to factors for a leak, use of a diverting ostomy is recommended.
the technical challenges associated with operating in the While it may not decrease the rate of anastomotic leak, it will
narrow male pelvis.22 This supposition is supported by the minimize the clinical impact of a leak on the patient.
findings that male gender is not associated with an increased
risk of leak following intraperitoneal anastomoses.23 A simi-
Operative Approach for Anastomotic Leaks
lar relationship between obesity and anastomotic leak has
been identified. One retrospective study found that obesity Anastomotic leaks should be treated based on their location,
was associated with higher risk of anastomotic leak following the stability of the patient, the degree of leakage (contained
low colorectal anastomoses.24 However, this association has vs. free perforation), and findings at the time of surgical
not been confirmed in larger, modern, but still retrospective, exploration. A modern algorithm for the treatment of intra-
studies including all anastomoses.20 It may be that the same peritoneal anastomotic leaks is presented in ►Fig. 1.
challenges that result in higher leak rates in the narrow, male If an anastomotic leak is suspected, assessing the stabi-
pelvis also exist in the pelvis of obese patients. lity of the patient is the first priority. If the patient is
Perioperative corticosteroids are often implicated as a risk hemodynamically unstable, immediate resuscitation with
factor for anastomotic leak. Individual retrospective studies intravenous (IV) fluids and broad-spectrum IV antibiotics
have shown varying associations between steroid use and should be undertaken. Invasive monitoring and transfer to
anastomotic leak. The largest systematic review, however, the intensive care unit may be necessary. After a period of
which included 12 studies, showed a higher rate of anasto- resuscitation, if the patient stabilizes, imaging such as a CT
motic leak in those patients who used corticosteroids pre- scan with IV, oral, and, if indicated, rectal contrast can be
operatively (6.8 vs. 3.3%).25 considered to further aid management. However, if the
Because of the frequent use of nonsteroidal anti-inflamma- patient remains unstable or has signs of diffuse peritonitis,
tory drugs (NSAIDs) as part of multimodality, nonnarcotic pain continued resuscitation with emergent operative explora-
management following colorectal surgery, their association tion is indicated.

Clinics in Colon and Rectal Surgery Vol. 32 No. 3/2019

 Operative Management of Anastomotic Leaks after Colorectal Surgery Saur, Paulson 193

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Fig. 1 Proposed algorithm for the management of intraperitoneal anastomotic leaks. (A) Determination of patient stability should be made and
resuscitation undertaken prior to diagnostic imaging. If patient remains unstable, they should follow “free perforation” portion of the algorithm
and proceed to surgical exploration. If the patient is stable or stabilizes with resuscitation, diagnostic imaging can be undertaken. (B) Versus
diagnostic laparoscopy, if applicable. (C) þ/ Mucous fistula, where appropriate. (D) Please refer to free perforation algorithm.

Most patients with anastomotic leak do not present with stomas were required in 14 patients (2.5% of total cohort, 18%
fulminant sepsis or hemodynamic instability; most present of those with anastomotic leaks).30
with signs and symptoms such as fever, leukocytosis, increasing Whether laparoscopic or open, safely and efficiently diag-
abdominal pain, and tachycardia without instability. In this nosing and treating the leak should be the priority in surgical
cohort of patients, imaging should be considered to evaluate exploration. If there is a large phlegmon and the bowel is matted,
more definitively for leak before intervention is undertaken. CT attempts to mobilize and explore the anastomosis may poten-
scan with contrast is the most common imaging modality tially create more injuries. In these cases, washout, drainage, and
utilized given its almost universal availability in a hospital proximal diversion are the preferred treatment modality.
setting and its ability to reliably demonstrate pathology such as If the anastomosis is safely accessible, evaluation of the
free air, perianastomotic air or fluid, and collections/abscesses. bowel viability and anastomotic integrity can be performed. If a
If imaging reveals a contained leak with either mild clearly delineated defect that is less than 1 cm can be identified
surrounding inflammation/phlegmon or an associated col- and the surrounding bowel is otherwise healthy without gross
lection that is less than 3 cm, bowel rest and broad-spectrum inflammation, primary repair of the anastomosis with prox-
antibiotics can be initiated with continued serial abdominal imal diversion can be considered. Otherwise, the anastomosis
and clinical exams to evaluate for signs of development of a should be taken down and re-created if the remaining bowel is
free perforation. If there is an organized abscess that is more healthy and nonedematous. Proximal diversion should be
than 3 cm large or multiple abscesses, percutaneous drainage strongly considered even if new anastomosis looks healthy
can be utilized in addition to antibiotics.29 and intact, particularly with more distal leaks.
If the patient fails conservative measures, surgical explora- If the bowel is unhealthy or edematous, creating a new
tion is warranted. Boyce et al evaluated 555 patients under- anastomosis is not indicated. An end stoma should be created
going laparoscopic low anterior resection with anastomosis with either a Hartmann’s pouch or a mucous fistula of the distal
within 10 cm of the anus and described their management of bowel. The level of a colorectal anastomosis would typically
the 44 patients who had an anastomotic leak. Sixteen of the preclude one from bringing up the rectum as a mucous fistula
patients who were diagnosed with a leak already had a in the event of anastomotic leak. If a colorectal anastomosis has
diverting stoma and only four of those patients required to be taken down, multiple drains should be left in the pelvis
reoperation. Twenty-four of the 28 patients, however, who and the rectum should be drained with a rectal tube.
were diagnosed with a leak who did not have a diverting stoma In the case of low colorectal, coloanal, or ileoanal anasto-
did require reoperation. Seventy-four per cent of the surgical moses, which do not already have a diverting stoma in place,
re-explorations were performed laparoscopically and typically repair of an anastomotic leak is almost never possible. Prox-
consisted of formation of ileostomy or colostomy. Permanent imal diversion, washout, and perianastomotic drainage are

Clinics in Colon and Rectal Surgery Vol. 32 No. 3/2019

194 Operative Management of Anastomotic Leaks after Colorectal Surgery Saur, Paulson

usually adequate to control sepsis. In patients who already 7 Platell C, Barwood N, Dorfmann G, Makin G. The incidence of
have been proximally diverted at the index operation, manage- anastomotic leaks in patients undergoing colorectal surgery.
ment depends on severity of leak, as discussed earlier. If a Colorectal Dis 2007;9(01):71–79
8 Sujatha-Bhaskar S, Jafari MD, Hanna M, et al. An endoscopic
perianastomotic abscess is present, drainage can be performed
mucosal grading system is predictive of leak in stapled rectal
via a percutaneous or transanal route. If transanal drainage is anastomoses. Surg Endosc 2018;32(04):1769–1775
appropriate, this can be performed in the operating room 9 Braunschmid T, Hartig N, Baumann L, Dauser B, Herbst F.
under general anesthesia or sedation and the abscess can be Influence of multiple stapler firings used for rectal division on
drained via the anastomosis with a drain (Penrose drain, colorectal anastomotic leak rate. Surg Endosc 2017;31(12):
5318–5326
typically) left in place, if needed. In patients who have a
10 Murray AC, Chiuzan C, Kiran RP. Risk of anastomotic leak after
subclinical leak such as those who have a leak identified
laparoscopic versus open colectomy. Surg Endosc 2016;30(12):
only on contrast enema study, CT scan should be performed 5275–5282
to evaluate for a collection. If a collection is present, it should be 11 Nelson H, Sargent DJ, Wieand HS, et al; Clinical Outcomes of
drained, if possible. If there is a contained leak without an Surgical Therapy Study Group. A comparison of laparoscopically
associated collection, expectant management can be per- assisted and open colectomy for colon cancer. N Engl J Med 2004;

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
350(20):2050–2059
formed with repeat imaging after a period of 3 to 6 weeks.
12 Guillou PJ, Quirke P, Thorpe H, et al; MRC CLASICC trial group.
Ultimately, the goal of any treatment intervention for Short-term endpoints of conventional versus laparoscopic-
anastomotic leak is to control and treat sepsis. This goal assisted surgery in patients with colorectal cancer (MRC CLASICC
should be kept in mind as the treatment plan is performed. trial): multicentre, randomised controlled trial. Lancet 2005;365
(9472):1718–1726
13 Kingham TP, Pachter HL. Colonic anastomotic leak: risk
Conclusion factors, diagnosis, and treatment. J Am Coll Surg 2009;208
(02):269–278
Anastomotic leak is an unfortunate reality in colorectal 14 Jafari MD, Wexner SD, Martz JE, et al. Perfusion assessment in
surgery. There are many risk factors described for anasto- laparoscopic left-sided/anterior resection (PILLAR II): a multi-
motic leak, although few have been definitively proven institutional study. J Am Coll Surg 2015;220(01):82–92.e1
consistently in high-quality studies. It is likely that these 15 Choy PY, Bissett IP, Docherty JG, Parry BR, Merrie A, Fitzgerald A.
Stapled versus handsewn methods for ileocolic anastomoses.
risk factors are cumulative and clinical judgment by the
Cochrane Database Syst Rev 2011;7(09):CD004320
surgeon is necessary in evaluating a patient’s individual
16 Neutzling CB, Lustosa SA, Proenca IM, da Silva EM, Matos D.
risk for leak. When modifiable factors are present, attempts Stapled versus handsewn methods for colorectal anastomosis
should be made to optimize the patient prior to proceeding surgery. Cochrane Database Syst Rev 2012;15(02):CD003144
with an operation. In the end, assessment of each individual 17 Damen N, Spilsbury K, Levitt M, et al. Anastomotic leaks in
patient’s risk for leak must be made at the time of the colorectal surgery. ANZ J Surg 2014;84(10):763–768
18 Gastinger I, Marusch F, Steinert R, Wolff S, Koeckerling F, Lippert
operation and decisions regarding anastomosis and use of
H; Working Group ‘Colon/Rectum Carcinoma’. Protective defunc-
proximal diversion should be made accordingly. Treatment tioning stoma in low anterior resection for rectal carcinoma. Br J
of anastomotic leak should start with determination of Surg 2005;92(09):1137–1142
patient stability followed by resuscitation and diagnostic 19 Hüser N, Michalski CW, Erkan M, et al. Systematic review and
imaging or operative exploration. Operative findings will meta-analysis of the role of defunctioning stoma in low rectal
dictate surgical approach with the goal of controlling sepsis cancer surgery. Ann Surg 2008;248(01):52–60
20 Parthasarathy M, Greensmith M, Bowers D, Groot-Wassink T. Risk
and stabilizing the patient.
factors for anastomotic leakage after colorectal resection: a retro-
spective analysis of 17 518 patients. Colorectal Dis 2017;19(03):
288–298
21 van Rooijen SJ, Huisman D, Stuijvenberg M, et al. Intraoperative
References modifiable risk factors of colorectal anastomotic leakage: why
1 McDermott FD, Heeney A, Kelly ME, Steele RJ, Carlson GL, Winter surgeons and anesthesiologists should act together. Int J Surg
DC. Systematic review of preoperative, intraoperative and post- 2016;36(Pt A):183–200
operative risk factors for colorectal anastomotic leaks. Br J Surg 22 Law WI, Chu KW, Ho JW, Chan CW. Risk factors for anastomotic
2015;102(05):462–479 leakage after low anterior resection with total mesorectal exci-
2 Adams K, Papagrigoriadis S. Little consensus in either definition or sion. Am J Surg 2000;179(02):92–96
diagnosis of a lower gastro-intestinal anastomotic leak amongst 23 Choi HK, Law WL, Ho JW. Leakage after resection and intraper-
colorectal surgeons. Int J Colorectal Dis 2013;28(07):967–971 itoneal anastomosis for colorectal malignancy: analysis of risk
3 Bruce J, Krukowski ZH, Al-Khairy G, Russell EM, Park KG. Systematic factors. Dis Colon Rectum 2006;49(11):1719–1725
review of the definition and measurement of anastomotic leak after 24 Rullier E, Laurent C, Garrelon JL, Michel P, Saric J, Parneix M. Risk
gastrointestinal surgery. Br J Surg 2001;88(09):1157–1168 factors for anastomotic leakage after resection of rectal cancer. Br
4 Hyman N, Manchester TL, Osler T, Burns B, Cataldo PA. Anasto- J Surg 1998;85(03):355–358
motic leaks after intestinal anastomosis: it’s later than you think. 25 Eriksen TF, Lassen CB, Gögenur I. Treatment with corticosteroids
Ann Surg 2007;245(02):254–258 and the risk of anastomotic leakage following lower gastrointest-
5 Gessler B, Eriksson O, Angenete E. Diagnosis, treatment, and inal surgery: a literature survey. Colorectal Dis 2014;16(05):
consequences of anastomotic leakage in colorectal surgery. Int J O154–O160
Colorectal Dis 2017;32(04):549–556 26 Smith SA, Roberts DJ, Lipson ME, Buie WD, MacLean AR. Post-
6 Walker KG, Bell SW, Rickard MJ, et al. Anastomotic leakage is operative nonsteroidal anti-inflammatory drug use and intestinal
predictive of diminished survival after potentially curative resec- anastomotic dehiscence: a systematic review and meta-analysis.
tion for colorectal cancer. Ann Surg 2004;240(02):255–259 Dis Colon Rectum 2016;59(11):1087–1097

Clinics in Colon and Rectal Surgery Vol. 32 No. 3/2019

 Operative Management of Anastomotic Leaks after Colorectal Surgery Saur, Paulson 195

27 Kotagal M, Hakkarainen TW, Simianu VV, Beck SJ, Alfonso-Cristan- 29 Siewert B, Tye G, Kruskal J, et al. Impact of CT-guided drainage in
cho R, Flum DR. Ketorolac use and postoperative complications in the treatment of diverticular abscesses: size matters. AJR Am J
gastrointestinal surgery. Ann Surg 2016;263(01):71–75 Roentgenol 2006;186(03):680–686
28 Saleh F, Jackson TD, Ambrosini L, et al. Perioperative nonselective 30 Boyce SA, Harris C, Stevenson A, Lumley J, Clark D. Management of
non-steroidal anti-inflammatory drugs are not associated with low colorectal anastomotic leakage in the laparoscopic era: more
anastomotic leakage after colorectal surgery. J Gastrointest Surg than a decade of experience. Dis Colon Rectum 2017;60(08):
2014;18(08):1398–1404 807–814

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Clinics in Colon and Rectal Surgery Vol. 32 No. 3/2019