Accepted Manuscript

Should patients with obstructing colorectal cancer have proximal diversion?

Chaya Shwaartz, MD, Adam C. Fields, MD, Jake G. Prigoff, MD, Jeffrey J. Aalberg,
BS, Celia M. Divino, MD, FACS

PII: S0002-9610(16)30501-3
DOI: 10.1016/j.amjsurg.2016.08.005
Reference: AJS 12048

To appear in: The American Journal of Surgery

Received Date: 9 April 2016

Revised Date: 5 August 2016
Accepted Date: 8 August 2016

Please cite this article as: Shwaartz C, Fields AC, Prigoff JG, Aalberg JJ, Divino CM, Should patients
with obstructing colorectal cancer have proximal diversion?, The American Journal of Surgery (2016),
doi: 10.1016/j.amjsurg.2016.08.005.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Should patients with obstructing colorectal cancer have proximal diversion?

Chaya Shwaartz MD, Adam C. Fields MD, Jake G. Prigoff MD, Jeffrey J. Aalberg BS, Celia

M. Divino MD, FACS

Department of Surgery, Division of General Surgery, Icahn School of Medicine at Mount

PT
Sinai

RI
Corresponding Author:

SC
Celia M. Divino, MD, FACS
Chief, Division of General Surgery
Department of Surgery, Icahn School of Medicine at Mount Sinai

U
1 Gustave L. Levy Place, Box 1041
AN
New York, NY, 10029
Phone: 212-241-5499, Fax: 212-410-0111
M

celia.divino@mountsinai.org
D
TE
C EP
AC

1
 ACCEPTED MANUSCRIPT

Abstract

Background: Up to 20% of patients with colorectal cancer present with obstruction. The

goal of this study was to compare the short-term outcomes of patients with obstructing colon

PT
cancer that underwent resection and primary anastomosis with or without proximal diversion.

RI
Methods: The ACS-NSQIP Procedure Targeted Colectomy databases from 2012-2014 were

reviewed. Patients undergoing colorectal resection with or without diverting ostomy for

SC
obstructing colorectal cancer were analyzed. Propensity-score matched cohorts of diverted

and non-diverted patients were created accounting for patient characteristics. The primary

U
outcomes were thirty-day mortality, postoperative complications, and readmission.
AN
Results: There were 2,323 patients (92%) with no proximal diversion and 204 patients (8%)
M

with proximal diversion. In univariate analysis, patients with colorectal resection with

diversion were significantly more likely to have any complication (p = 0.001), sepsis (p =
D

0.01), and blood transfusion (p = 0.001). Diversion patients were also significantly more
TE

likely to be readmitted to the hospital within 30 days of the index procedure (p = 0.02).

Proximal diversion was associated with any complication (p = 0.01), failure to wean off
EP

ventilator (p = 0.05), and longer length of stay (p = 0.01) in matched cohorts.

C

(Table 2).
AC

Conclusions: Proximal diversion in the setting of obstructive colorectal cancer is associated

with higher rates of any complication, deep wound infection, sepsis, and readmission.

Surgeons who perform a primary anastomosis with diversion for obstructing colorectal

cancer should take into account the significant risk for postoperative complications.

Key words: Colorectal cancer; obstruction; diversion; stoma; complications

2
 ACCEPTED MANUSCRIPT

Introduction:

Colorectal cancer is the third most common cancer and the third leading cause of

cancer death in United States (1). Despite the improvement in diagnostic modalities and

PT
screening protocols, approximately 20% of the patients present with obstruction, mostly from

RI
tumors on the left side (2-4). The treatment of choice for obstructing colorectal cancer

depends on the general condition of the patient, the location of the tumor, and the degree of

SC
obstruction. Various palliative and curative procedures may be considered. However, primary

resection of the tumor is the preferred option for the patient when possible (2, 5-9). In

U
patients who undergo resection, there is still a debate whether to perform a one-stage or
AN
multi-stage operation. Traditionally, patients with left sided obstructing colorectal cancer
M

were treated with a Hartman’s procedure (3, 10, 11). In the past decade, resection and

primary anastomosis gained popularity over Hartman’s procedure in low risk patients (2, 4, 5,
D

12). Often, primary anastomosis is protected by a diverting ileostomy in order to prevent the
TE

morbidity and mortality associated with anastomotic leak (13-15). However, it has been

shown that the morbidity of ileostomy creation and its closure may reach 50% and includes
EP

anastomotic leak, bowel obstruction, surgical site infection, parastomal hernia, dehydration
C

and readmissions (16-18).

AC

There is some evidence to suggest that primary anastomosis without proximal

diversion in patients with obstructing colorectal cancer may lead to a lower rate of

postoperative complications (7, 19-24). However, large-scale studies are lacking and there is

still a debate as to whether or not proximal diversion should be routine in those patients or

should be avoided when possible.

3
 ACCEPTED MANUSCRIPT
The aim of this study is to compare the outcomes of patients with obstructing colon

cancer that underwent resection and primary anastomosis with or without proximal diversion.

This study was carried out using the American College of Surgeons’ National Surgical

Quality Improvement Program (ACS NSQIP) a large, multicenter, prospectively-collected

database.

PT
RI
U SC
AN
M
D
TE
C EP
AC

4
 ACCEPTED MANUSCRIPT

Materials and Methods

Data Collection

PT
The 2012, 2013, and 2014 NSQIP Procedure-Targeted Colectomy Databases were

RI
used in this study. The Procedure-Targeted Colectomy Database is designed for high-risk,

high-volume procedures for large or specialty hospitals and includes 22 perioperative

SC
variables specific for colorectal procedures including bowel prep, anastomotic leak, and

postoperative ileus. Many of these variables are specific for colon cancer including

U
chemotherapy, resection margins, and cancer staging. Multicenter prospective data were
AN
collected from 203 hospitals (25).
M

Patients were coded in NSQIP as having a surgical indication of colorectal cancer

with obstruction. This was defined based on the surgeon’s postoperative diagnosis and/or the
D

pathology reports. Patients included in the study underwent a colorectal resection with
TE

primary anastomosis defined by Current Procedural Terminology (CPT) codes of the

American Medical Association 44140, 44145, 44146, 44160, 44204, 44205, 44207 and
EP

44208. Stoma construction was identified via CPT codes 44310 and 44320 There were 2,527
C

eligible patients.
AC

The Procedure-Targeted Colectomy Database and traditional NSQIP databases

include over 300 common perioperative variables. Data includes demographics,

comorbidities, postoperative outcomes up to 30 days (data beyond 30 days are not available),

and other variables. Outcome variables include mortality, need for reoperation, duration of

stay, and in-hospital and out-of-hospital complications. Access to the NSQIP database is

available to all investigators at ACS-NSQIP participating hospitals.

5
 ACCEPTED MANUSCRIPT
Outcomes

Primary outcomes in this study are mortality, postoperative complications, and

hospital readmission. Major gastrointestinal complications specific to the Procedure-Targeted

Colectomy Database include anastomotic leak and prolonged postoperative NPO or NGT

use. Non-gastrointestinal specific postoperative complications include wound infection,

PT
cardiac arrest requiring CPR, myocardial infarction, septic shock, sepsis, coma, stroke,

RI
urinary tract infection, acute renal failure, renal insufficiency, pneumonia, reintubation,

failure to wean from ventilator within 48 hours, blood transfusion, deep vein thrombosis,

SC
pulmonary embolism, and reoperation. These outcomes are assessed in-hospital and out-of-

hospital for 30 days.

Statistical analysis
U
AN
Univariate analysis was conducted using Pearson chi-square and Fisher’s exact tests
M

for categorical variables and independent t-tests for normally distributed continuous

variables. Significance was defined as P < 0.05 for all tests.

D

Rates of baseline co-morbidities and patient’s characteristics differed substantially

TE

between patients that underwent diversion to patients that did not undergone diversion. In

order to control for these significant differences between the two patient populations studied
EP

(no diversion versus diversion patients), propensity score matching was utilized. The use of
C

propensity score matching has been shown to reduce bias (26, 27). A logistic regression
AC

model was fitted with diversion vs. non-diversion as outcome and age, race, and co-

morbidities as co-variates. The model’s discrimination, based on a C statistic, was 0.82. The

patients were matched using a ‘optimally’ matching algorithm. Table 1 compares the baseline

characteristics of the diverted patients and the non-diverted patients matched by this

algorithm. Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary,

NC).

6
 ACCEPTED MANUSCRIPT

Results:

There were a total of 2,527 patients who underwent colorectal resection with

PT
anastomosis with or without proximal diversion. There were 2,323 patients (92%) with no

RI
proximal diversion and 204 patients (8%) with proximal diversion. The demographics and

clinical characteristics of the patients are described in Table 1. There were several factors

SC
between the two groups that were significantly different from one another including age,

medical comorbidities, and procedure characteristics. To account for these differences,

U
propensity score matching was done. These differences were not seen between propensity-
AN
matched cohorts. In multivariate logistic regression analysis, predictive factors for
M

undergoing diversion included recent weight loss (p = 0.03), higher wound class (p <

0.0001), chemotherapy within 90 days (p = 0.02), low pelvic procedures (p < 0.001), and
D

open procedures (p < 0.001).

TE

In the unmatched cohort there was a total of 1,261 patients (49.9%) with

postoperative complications within thirty days of surgery (Table 2). Patients with colorectal
EP

resection with diversion were significantly more likely to have any complication (61.2%
C

versus 48.9%, p = 0.001), sepsis (8.8% versus 4.8%, p = 0.01), and postoperative bleeding
AC

(25% versus 15.2%, p = 0.001). Diversion patients were also significantly more likely to be

readmitted to the hospital within 30 days of the index procedure (15.2% versus 9.9%, p =

0.02). There was no difference in mortality (4.9% versus 5.2%, p = 0.87), anastomotic leak

rate (3.4% versus 4.4%, p = 0.49), or ileus (27.9% versus 23.1%, p = 0.12) between the two

groups.

7
 ACCEPTED MANUSCRIPT
Proximal diversion was associated with any complication (p = 0.01), failure to wean

off ventilator (p = 0.05), and longer length of stay (p = 0.01) in matched cohorts. However,

sepsis and readmission rates were similar after propensity score matching (p = 0.2, p = 0.15).

Discussion:

PT
This study found that patients who underwent primary anastomosis with proximal

RI
diversion have significantly higher rates of negative outcomes such as sepsis, longer length of

stay and 30-day readmission.

SC
In many cases, the location of the tumor is an important factor in the decision of the

preferred procedure. It is generally accepted that when the tumor is located proximal to the

U
splenic flexure, the preferred operation is an extended right hemicolectomy with ileocolic
AN
anastomosis unless the patient does not fit to have anastomosis (2, 5, 11, 12, 28, 29).
M

However, when the obstructing tumor is located distal to the splenic flexure there are number

of options for the operating surgeon (2-5, 10, 12, 19, 29-32). The traditional procedure for
D

left sided obstructing tumor was Hartman’s procedure with colostomy, which may be
TE

reversed in an additional major surgery. Another option is resection and primary anastomosis

with or without proximal diversion. In this study, we focused on the patients that had primary
EP

anastomosis and excluded patients that underwent either loop colostomy without resection or
C

Hartman’s procedure. Most of the patients in the diversion group had low pelvic procedures;
AC

it is commonly believed that diversion in these cases improves outcomes (13, 33, 34). In

2008, a meta-analysis conducted by Hüser et al. evaluating the role of diverting stoma in

patients undergoing low anterior resection for rectal cancer mostly in the elective setting

showed that there was no difference in the rate of anastomotic leaks or mortality rate between

the two groups. However, the risk for reoperation in the patients that did not have diverting

stoma was significantly higher and the authors concluded that defunctioning stoma is

8
 ACCEPTED MANUSCRIPT
recommended in those patients and reduces the rate of clinically relevant anastomotic leaks

(33). On the other hand, there were multiple studies that showed no advantage for diverting

stoma in those patients undergoing pelvic surgery (12, 19, 35) . In 2013, Nurkin et al.

retrospectively review 1,791 patients that underwent low anterior resection from multiple

hospitals. Six hundred and six patients received diverting stoma and 660 patients underwent

PT
low anterior resection without proximal diversion. They found no differences in wound

RI
complications, sepsis, or incidence of septic shock between the two groups. Furthermore,

patients in the stoma group were more likely to develop postoperative acute kidney injury

SC
(35).

Emergency operations in patients with obstructing colon cancer are associated with

U
high morbidity and mortality despite advanced perioperative care; thus, a diverting stoma
AN
may specifically be indicated in the setting of obstructing tumor (6, 11, 36). These patients
M

are often underresuscitated and did not receive bowel prep, and the bowel proximal to the

obstruction site is often dilated. In the 1990s, colonic stenting was introduced in order to
D

restore luminal patency in patients with obstructing colon cancer (37). Multiple studies have
TE

shown that stent placement as a bridge to surgery, can improve the clinical condition of the

patient and decrease mortality, morbidity, and number of colostomies (9, 20, 37-39). Our data
EP

showed a high incidence of primary anastomosis without proximal diversion in patients with
C

obstructing colorectal cancer. Only 8.0% of patients in our study underwent proximal
AC

diversion. The use of self-expanding metal stents is not provided in the ACS-NSQIP and may

be one of the reasons for avoidance of proximal stoma.

The rate of complications in our cohort was 50%. Kube et al. had a total complication

rate of 30%. They compared three groups of patients with obstructing colorectal cancer

(primary anastomosis without proximal diversion, primary anastomosis with proximal

diversion and Hartman’s procedure). The authors included patients with Hartman’s procedure

9
 ACCEPTED MANUSCRIPT
which is usually associated with higher rate of complications. However, there was no

significant difference in the rate of complications between the groups. These results

demonstrate that there is no advantage to the use of proximal diversion over primary

anastomosis without proximal diversion (8). Another study by Anderin et al. compared

patients that underwent low anterior resection with diverting ostomy to patients that

PT
underwent low anterior resection without a proximal ostomy. They reported an overall

RI
postoperative complication rate similar to the rate in our study (48%) and there was no

significant difference in complication rate between the two groups (40).

SC
In our study, diverting ostomy was associated with higher risk of deep wound

infection and sepsis in the non-matched cohorts. A recent study showed that proximal

U
diversion was associated with deep wound infection and sepsis/septic shock in patients that
AN
underwent elective colectomy for diverticular disease (41). However, in the matched cohorts,
M

the rates of these outcomes were similar.

Ostomy was an independent predictor of longer length of stay and hospital

D

readmission within 30 days of index procedure. There are several studies suggesting that the
TE

complication rate of ostomy is up to 50% including high output stoma, dehydration, renal

failure, parastomal hernia and small bowel obstruction (16, 17, 41-43). Additionally, the data
EP

regarding the rate of complications of temporary ileostomy more than 30 days from surgery
C

are lacking. Furthermore, approximately 25% of these patients will not have their ostomy
AC

reversed (42, 43).

The results of this study suggest that proximal diversion in patients undergoing

resection with primary anastomosis for obstructing colorectal cancer has worse outcomes

compared to patients that did not have diversion. There are many strengths to this study.

First, this is one of the largest studies with patients from multiple institutions to assess

outcomes in patients with obstructing colon cancer. Second, we were able to use procedure

10
 ACCEPTED MANUSCRIPT
targeted data, which includes variables specific to colorectal procedures. Third, the data

recorded in NSQIP are continually validated by reviewers.

However, there are potential limitations and biases that need to be addressed. The

main drawback of this study is its retrospective nature without complete information for some

variables collected. Given the retrospective nature of data collection, we are unable to

PT
determine what factors led the surgeon to divert the patient. One possible bias is that patients

RI
in better clinical states or patients who had better expected prognoses were more likely to

have primary anastomosis without proximal diversion. Further, it is possible that the rate of

SC
complications would have been even higher if the patients with diversion had not had such a

procedure. It is also possible that some patients with diversion were missed from miscoding.

U
Additionally, some of the variables are not recorded in the database such as intraoperative
AN
findings that may have led the surgeon to divert the patient, the use of metal stents as a bridge
M

to surgery, and the use of intraoperative lavage. There may also be variations between sites in

the definition of obstructing tumor, which may potentially be used for endoscopically
D

obstructing tumors, clinical complete obstruction, and obstructions that were alleviated by the
TE

use of stents.

Conclusions:
EP

The results of this study suggest that proximal diversion in the setting of obstructive
C

colorectal cancer may be associated with increased morbidity, specifically higher rates of any
AC

complication and longer length of stay. The management of patients with obstructing

colorectal cancer should be individualized based on patient factors, experience of the

surgeon, and intraoperative findings. However, when performing a primary anastomosis with

diversion one should take into account the significant risk for complications. Randomized

prospective trials are required to define the role of diverting stoma in patients undergoing

resection and primary anastomosis for obstructing colorectal cancer.

11
 ACCEPTED MANUSCRIPT

PT
References:

RI
1. Siegel R, DeSantis C, Jemal A. Colorectal cancer statistics, 2014. CA Cancer J Clin,
2014;64:104-17.
2. Schwenter F, Morel P, Gervaz P. Management of obstructive and perforated

SC
colorectal cancer. Expert Rev Anticancer Ther, 2010;10:1613-9.
3. Deans G, Krukowski Z, Irwin S. Malignant obstruction of the left colon. Br J Surg,
1994;81:1270-6.
4. Danis J. Single-stage treatment for malignant left-sided colonic obstruction: a

U
prospective randomized clinical trial comparing subtotal colectomy with segmental
AN
resection following intraoperative irrigation. Br J Surg, 1996;83:1303.
5. Lee-Kong S, Lisle D. Surgical Management of Complicated Colon Cancer. Clin Colon
Rectal Surg, 2015;28:228-33.
6. Smothers L, Hynan L, Fleming J, et al. Emergency surgery for colon carcinoma. Dis
M

Colon Rectum, 2003;46:24-30.

7. Zorcolo L, Covotta L, Carlomagno N, Bartolo DC. Safety of primary anastomosis in
emergency colo-rectal surgery. Colorectal Dis, 2003;5:262-9.
D

8. Kube R, Granowski D, Stubs P, et al. Surgical practices for malignant left colonic
obstruction in Germany. Eur J Surg Oncol, 2010;36:65-71.
TE

9. Ansaloni L, Andersson RE, Bazzoli F, et al. Guidelenines in the management of

obstructing cancer of the left colon: consensus conference of the world society of
emergency surgery (WSES) and peritoneum and surgery (PnS) society. World J Emerg Surg,
EP

2010;5:29.
10. Chereau N, Lefevre JH, Lefrancois M, et al. Management of malignant left colonic
obstruction: is an initial temporary colostomy followed by surgical resection a better
C

option? Colorectal Dis, 2013;15:e646-53.

11. Saliangas K, Economou A, Nikoloudis N, et al. Treatment of complicated colorectal
AC

cancer. Evaluation of the outcome. Tech Coloproctol, 2004;8 Suppl 1:s199-201.

12. Hsu TC. Comparison of one-stage resection and anastomosis of acute complete
obstruction of left and right colon. Am J Surg, 2005;189:384-7.
13. Thoker M, Wani I, Parray FQ, et al. Role of diversion ileostomy in low rectal cancer: a
randomized controlled trial. Int J Surg, 2014;12:945-51.
14. Gastinger I, Marusch F, Steinert R, et al. Protective defunctioning stoma in low
anterior resection for rectal carcinoma. Br J Surg, 2005;92:1137-42.
15. Fasth S, Hulten L. Loop ileostomy: a superior diverting stoma in colorectal surgery.
World J Surg, 1984;8:401-7.
16. Thalheimer A, Bueter M, Kortuem M, et al. Morbidity of temporary loop ileostomy in
patients with colorectal cancer. Dis Colon Rectum, 2006;49:1011-7.

12
 ACCEPTED MANUSCRIPT
17. Luglio G, Pendlimari R, Holubar SD, et al. Loop ileostomy reversal after colon and
rectal surgery: a single institutional 5-year experience in 944 patients. Arch Surg,
2011;146:1191-6.
18. Kaidar-Person O, Person B, Wexner SD. Complications of construction and closure of
temporary loop ileostomy. J Am Coll Surg, 2005;201:759-73.
19. Breitenstein S, Rickenbacher A, Berdajs D, et al. Systematic evaluation of surgical
strategies for acute malignant left-sided colonic obstruction. Br J Surg, 2007;94:1451-60.
20. Park IJ, Choi GS, Kang BM, et al. Comparison of one-stage managements of
obstructing left-sided colon and rectal cancer: stent-laparoscopic approach vs.

PT
intraoperative colonic lavage. J Gastrointest Surg, 2009;13:960-5.
21. Pisanu A, Montisci A, Cillara N, et al. One-stage treatment in emergency setting for
obstructing colorectal cancer. Suppl Tumori, 2005;4:S148.

RI
22. Stipa F, Pigazzi A, Bascone B, et al. Management of obstructive colorectal cancer
with endoscopic stenting followed by single-stage surgery: open or laparoscopic resection?

SC
Surg Endosc, 2008;22:1477-81.
23. Baccari P, Bisagni P, Crippa S, et al. Operative and long-term results after one-stage
surgery for obstructing colonic cancer. Hepatogastroenterology, 2006;53:698-701.
24. Shigeta K, Baba H, Yamafuji K, et al. Outcomes for patients with obstructing

U
colorectal cancers treated with one-stage surgery using transanal drainage tubes. J
Gastrointest Surg, 2014;18:1507-13.
AN
25. American College of Surgeons. ACS NSQIP Participant Use Data File
[cited July 27, 2016]; Available from: https://www.facs.org/quality-programs/acs-
nsqip/program-specifics/participant-use.
M

26. D'Agostino RB, Jr. Propensity score methods for bias reduction in the comparison of
a treatment to a non-randomized control group. Stat Med, 1998;17:2265-81.
27. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of
D

Confounding in Observational Studies. Multivariate Behav Res, 2011;46:399-424.

28. Formisano V, Di Muria A, Connola G, et al. Our experience in the management of
TE

obstructing colorectal cancer. Ann Ital Chir, 2014;85:563-8.

29. Lee YM, Law WL, Chu KW, Poon RT. Emergency surgery for obstructing colorectal
cancers: a comparison between right-sided and left-sided lesions. J Am Coll Surg,
EP

2001;192:719-25.
30. Amelung FJ, Mulder CL, Verheijen PM, et al. Acute resection versus bridge to surgery
with diverting colostomy for patients with acute malignant left sided colonic obstruction:
C

Systematic review and meta-analysis. Surg Oncol, 2015;24:313-21.

31. Yang HY, Wu CC, Jao SW, et al. Two-stage resection for malignant colonic
AC

obstructions: the timing of early resection and possible predictive factors. World J
Gastroenterol, 2012;18:3267-71.
32. Wang JH, Kung YH, King TM, et al. Comparison of laparoscopic versus open surgery in
a three-stage operation for obstructive left-sided colorectal cancer. J Chin Med Assoc,
2015;78:584-90.
33. Huser N, Michalski CW, Erkan M, et al. Systematic review and meta-analysis of the
role of defunctioning stoma in low rectal cancer surgery. Ann Surg, 2008;248:52-60.
34. Hanna MH, Vinci A, Pigazzi A. Diverting ileostomy in colorectal surgery: when is it
necessary? Langenbecks Arch Surg, 2015;400:145-52.

13
 ACCEPTED MANUSCRIPT
35. Nurkin S, Kakarla VR, Ruiz DE, et al. The role of faecal diversion in low rectal cancer: a
review of 1791 patients having rectal resection with anastomosis for cancer, with and
without a proximal stoma. Colorectal Dis, 2013;15:e309-16.
36. Ng HJ, Yule M, Twoon M, et al. Current outcomes of emergency large bowel surgery.
Ann R Coll Surg Engl, 2015;97:151-6.
37. Martinez-Santos C, Lobato RF, Fradejas JM, et al. Self-expandable stent before
elective surgery vs. emergency surgery for the treatment of malignant colorectal
obstructions: comparison of primary anastomosis and morbidity rates. Dis Colon Rectum,
2002;45:401-6.

PT
38. Saida Y, Sumiyama Y, Nagao J, Uramatsu M. Long-term prognosis of preoperative
"bridge to surgery" expandable metallic stent insertion for obstructive colorectal cancer:
comparison with emergency operation. Dis Colon Rectum, 2003;46:S44-9.

RI
39. Ng KC, Law WL, Lee YM, et al. Self-expanding metallic stent as a bridge to surgery
versus emergency resection for obstructing left-sided colorectal cancer: a case-matched

SC
study. J Gastrointest Surg, 2006;10:798-803.
40. Anderin K, Gustafsson UO, Thorell A, Nygren J. The effect of diverting stoma on
postoperative morbidity after low anterior resection for rectal cancer in patients treated
within an ERAS program. Eur J Surg Oncol, 2015;41:724-30.

U
41. Wise KB, Merchea A, Cima RR, et al. Proximal intestinal diversion is associated with
increased morbidity in patients undergoing elective colectomy for diverticular disease: an
AN
ACS-NSQIP study. J Gastrointest Surg, 2015;19:535-42.
42. Chun LJ, Haigh PI, Tam MS, Abbas MA. Defunctioning loop ileostomy for pelvic
anastomoses: predictors of morbidity and nonclosure. Dis Colon Rectum, 2012;55:167-74.
M

43. Jafari MD, Halabi WJ, Jafari F, et al. Morbidity of diverting ileostomy for rectal
cancer: analysis of the American College of Surgeons National Surgical Quality Improvement
Program. Am Surg, 2013;79:1034-9.
D
TE
C EP
AC

14
 ACCEPTED MANUSCRIPT

Table 1. Patient and procedure characteristics

PT
Factor Before Propensity Matching After Propensity Matching

RI
No Diversion Diversion P value No Diversion Diversion P value
(N = 2,323) (N = 204) (N = 179) (N = 179)

SC
Demographics

U
Age 68.4 ± 14.9 65.2 ± 14.7 0.004 64.2 ± 15.2 65.1 ± 14.5 0.57

AN
Female gender 1,095 (47.1%) 100 (49.0%) 0.61 83 (46.4%) 84 (46.9%) 0.92

M
Race 0.81 0.82

D
White 1,604 (69.0%) 142 (69.6%) 127 (71.0%) 132 (73.4%)

Black
Other
250 (10.8%)
469 (20.2%)
24 (11.8%)
38 (18.6%) TE 21 (11.7%)
31 (17.3%)
20 (11.2%)
27 (15.1%)
EP
Clinical
characteristics
C

ASA 1 or 2 734 (31.6%) 63 (30.9%) 0.86 60 (33.5%) 56 (31.5%) 0.68

AC

Functional status 2,190 (94.3%) 192 (94.1%) 0.91 164 (91.6%) 168 (93.4%) 0.42
independent

BMI, kg/m2 26.7 ± 6.4 26.5 ± 6.7 0.67 26.5 ± 6.8 26.6 ± 6.7 0.87

1
 ACCEPTED MANUSCRIPT

Smoker 371 (16.0%) 48 (23.5%) 0.005 41 (22.9%) 43 (24.0%) 0.80

Steroid use 73 (3.1%) 8 (4.0%) 0.55 6 (3.4%) 8 (4.5%) 0.59

PT
Weight loss 317 (13.6%) 44 (21.6%) 0.002 43 (24.0%) 38 (21.2%) 0.53

RI
Wound class 3 or 4 418 (18.0%) 67 (32.8%) <0.001 55 (30.7%) 61 (34.1%) 0.50

SC
Disseminated 547 (23.5%) 65 (31.9%) 0.008 52 (29.1%) 56 (31.3%) 0.65
Cancer

U
Chemotherapy 115 (5.0%) 26 (12.7%) <0.001 19 (10.6%) 22 (12.3%) 0.62

AN
within 90 days

M
Lab Values

Hematocrit 34.8 ± 6.0 35.1 ± 6.0 0.44 35.5 ± 5.7 35.1 ± 5.6 0.59

D
TE
Creatinine 0.9 ± 0.6 1.0 ± 0.7 0.48 0.86 ± 0.3 0.98 ± 0.8 0.07

Albumin 3.5 ± 0.7 3.4 ± 0.7 0.14 3.4 ± 0.73 3.3 ± 0.72 0.16
EP
Tumor staging
C

T stage, T3/T4 1,966 (84.6%) 181 (88.7%) 0.21 155 (95.1%) 160 (96.4%) 0.56
AC

N stage, N1/N2 1,280 (55.1%) 115 (56.4%) 0.80 100 (62.1%) 104 (64.6%) 0.64

M stage, M1a 383 (31.0%) 48 (37.8%) 0.12 30 (30.3%) 42 (35.9%) 0.38

2
 ACCEPTED MANUSCRIPT

Comorbidities

COPD 162 (7.0%) 6 (3.0%) 0.03 5 (2.8%) 6 (3.4%) 0.76

PT
CHF 38 (1.6%) 2 (0.9%) 0.77 2 (1.1%) 2 (1.1%) 1.00

RI
Hypertension 1,205 (51.9%) 106 (52.0%) 0.98 87 (48.6%) 91 (50.8%) 0.67

SC
Diabetes 363 (15.6%) 32 (15.7%) 0.98 21 (11.7%) 29 (16.2%) 0.22

Procedure

U
Characteristics

AN
Procedure <0.001 0.9

M
Ileocolic resection 800 (34.4%) 37 (18.1%) 35 (19.6%) 32 (17.9%)

D
Partial colectomy 1,170 (50.4%) 41 (20.1%) 32 (17.9%) 35 (19.6%)

TE
Low pelvic 353 (15.2%) 126 (61.8%) 112 (62.6%) 112 (62.6%)
EP
Operative approach <0.001 0.36

Open 1,484 (63.9%) 164 (80.4%) 138 (77.1%) 145 (81.0%)

C
AC

Minimally 828 (35.6%) 40 (19.6%) 41 (22.9%) 34 (19.0%)

invasive

Mechanical bowel 888 (38.2%) 71 (34.8%) 0.15 74 (41.3%) 70 (39.1%) 0.67

prep

3
 ACCEPTED MANUSCRIPT

Oral antibiotics 367 (15.8%) 34 (16.7%) 0.85 34 (19.5%) 34 (19.5%) 0.98

Emergency 641 (27.6%) 54 (26.5%) 0.73 38 (21.2%) 43 (24.0%) 0.53

PT
Emergency 0.04 0.63
indication

RI
Obstruction 588 (25.3%) 45 (22.1%) 33 (18.5%) 37 (20.7%)

SC
Perforation 31 (1.3%) 8 (3.9%) 2 (1.1%) 5 (2.8%)

U
Bleeding 10 (0.4%) 0 (0%) 1 (0.6%) 0 (0%)

AN
Other 9 (0.4%) 1 (0.5%) 1 (0.6%) 1 (0.6%)

M
Operative time, min 153 ± 79 180 ± 110 0.001 157.4 ± 76.8 180.7 ± 112.1 0.02

D
TE
EP
Continuous variables given as mean ± SD, categorical variables given as
percentages
ASA, American society of anesthesia; BMI, body mass index; COPD, chronic
C

obstructive
AC

pulmonary disease
a
Based on available data

4
 ACCEPTED MANUSCRIPT

Table 2. Thirty-day postoperative complications

PT
Factor Before Propensity Matching After Propensity Matching

RI
No Diversion Diversion P value No Diversion Diversion P value
(N = 2,323) (N = 204) (N = 179) (N = 179)

SC
Any complication 1,136 (48.9%) 125 0.001 85 (47.5%) 108 (60.3%) 0.01
(61.2%)

U
Superficial wound 171 (7.4%) 16 (7.8%) 0.80 15 (8.4%) 13 (7.3%) 0.69

AN
infection

Deep wound 47 (2.0%) 8 (3.9%) 0.08 5 (2.8%) 7 (3.9%) 0.56

M
infection

D
Organ space 121 (5.2%) 8 (3.9%) 0.42 7 (3.9%) 8 (4.5%) 0.79

TE
infection

Wound dehiscence 40 (1.7%) 3 (1.5%) 1.00 4 (2.2%) 2 (1.1%) 0.41

EP
Pneumonia 102 (4.4%) 11 (5.4%) 0.51 7 (3.9%) 9 (5.0%) 0.61
C

Unplanned 84 (3.6%) 6 (3.0%) 0.62 4 (2.2%) 6 (3.4%) 0.52

AC

intubation

Deep-vein 53 (2.3%) 5 (2.5%) 0.81 2 (1.1%) 5 (2.8%) 0.25

thrombosis

Failure to wean 65 (2.8%) 8 (3.9%) 0.36 2 (1.1%) 8 (4.5%) 0.05

5
 ACCEPTED MANUSCRIPT

from ventilator

Pulmonary 28 (1.2%) 0 (0%) 0.16 2 (1.1%) 0 (0%) 0.16

embolism

PT
Renal insufficiency 18 (0.8%) 2 (1.0%) 0.67 1 (0.6%) 2 (1.1%) 0.56

RI
Acute renal failure 17 (0.7%) 2 (1.0%) 0.66 0 (0%) 2 (1.1%) 0.16

SC
Urinary tract 86 (3.7%) 9 (4.4%) 0.61 8 (4.5%) 7 (3.9%) 0.79
infection

U
Stroke 11 (0.5%) 2 (1.0%) 0.28 1 (0.6%) 2 (1.1%) 0.56

AN
Cardiac arrest 20 (0.9%) 4 (2.0%) 0.12 2 (1.1%) 4 (2.2%) 0.41

M
Myocardial 27 (1.2%) 2 (1.0%) 1.00 2 (1.1%) 1 (0.6%) 0.56

D
infarction

TE
Bleeding 370 (15.2%) 51 (25.0%) 0.001 32 (17.9%) 43 (24.0%) 0.15

Sepsis 111 (4.8%) 18 (8.8%) 0.01 9 (5.0%) 15 (8.4%) 0.20

EP
Septic shock 79 (3.4%) 10 (4.9%) 0.27 6 (3.4%) 9 (5.0%) 0.43
C

Ileus 537 (23.1%) 57 (27.9%) 0.12 42 (23.6%) 49 (27.4%) 0.41

AC

Anastomotic Leak 103 (4.4%) 7 (3.4%) 0.49 9 (5.0%) 6 (3.4%) 0.43

Reoperation 148 (6.4%) 9 (4.4%) 0.27 13 (7.3%) 7 (3.9%) 0.17

6
 ACCEPTED MANUSCRIPT

Readmission 229 (9.9%) 31 (15.2%) 0.02 18 (10.1%) 27 (15.1%) 0.15

PT
Mortality 120 (5.2%) 10 (4.9%) 0.87 12 (6.7%) 8 (4.5%) 0.36

RI
LOS, days 10.7 ± 9.2 12.0 ± 8.9 0.06 10.1 ± 7.2 12.3 ± 9.2 0.01

SC
LOS, length of stay

U
AN
M
D
TE
C EP
AC