Original article
doi:10.1111/j.1463-1318.2005.00875.x
Management of the rectal stump after emergency sub-total
colectomy: which surgical option is associated with the lowest
morbidity?
J. P. Trickett, H. S. Tilney, A. M. Gudgeon, S. G. Mellor and D. P. Edwards
Department of Colorectal Surgery, Frimley Park Hospital, Frimley, UK
Received 19 April 2005; accepted 26 April 2005
Abstract
Objective To identify the preferred surgical management of the rectal stump after emergency subtotal
colectomy (ESC) for acute severe colitis by assessing the
morbidity associated with each option.
Patients and methods Consecutive patients undergoing ESC at a district general hospital between 1999
and 2004 were retrospectively audited for pathology,
rectal stump complications and length of postoperative
hospital stay (POS).
Results Thirty-seven ESCs were performed, 34 were
undertaken for disease refractory to medical treatment, 2
for toxic mega colon and 1 for perforation. Thirty-four
cases were for ulcerative colitis, 2 Crohn’s colitis and 1
infective colitis. Twenty-seven had an intraperitoneal and
10 a subcutaneously placed closed rectal stump. The
median POS for patients with a subcutaneously placed
stump was shorter than for those with an intraperitoneal
Introduction
Subtotal colectomy and ileostomy with preservation of
the rectal stump is established as the preferred operation
for acute severe colitis which fails to respond to medical
therapy [1–3]. The surgical management of the rectal
stump, however, remains controversial. The options
include creation of a low sigmoid mucous fistula, closure
of the rectosigmoid but leaving the closed stump in the
subcutaneous plane at the lower end of a midline wound,
or closure of the rectal stump at the level of the sacral
promontory (leaving the rectal suture ⁄ staple-line in the
peritoneal cavity). The former options necessitate a
Presented as a poster at the European Association of Coloproctology, Sitges,
Spain. September 2003
Correspondence to: Mr D. P. Edwards, Department of Colorectal Surgery, Frimley
Park Hospital, Frimley, GU16 7UJ, UK.
E-mail: david.edwards@fph-tr.nhs.uk
2005 Blackwell Publishing Ltd. Colorectal Disease, 7, 519–522
stump, 8 and 15 days, respectively (P ¼ 0.04). Two
patients had leakage from an intraperitoneal stump,
prolonging POS (33 and 193 days). Three of the
subcutaneous stumps leaked causing wound infection
but not prolonging the POS (6, 7 and 16 days).
Conclusion Avoiding a second stoma by closing the
rectal stump after ESC has been confirmed as acceptable practice by studies over the last 15 years, reporting
no overall increase in complications. The location of a
closed rectal stump appears to influence the incidence
of pelvic sepsis. The lowest pelvic sepsis rate is
associated with subcutaneous placement; despite a
higher wound infection rate this option appears to be
associated with a lower total morbidity reflected in a
shorter POS.
Keywords Emergency sub-total colectomy (ESC), morbidity, rectal stump, surgical management
longer stump and therefore greater risk of complications
from the retained inflamed bowel; however, the latter
option may result in an intraperitoneal leak from the
closed rectal stump.
Studies over the last 15 years have suggested that
avoidance of a mucous fistula is not associated with an
overall increase in the risk of pelvic sepsis (Table 1). In
order to identify the preferred surgical option for a closed
rectal stump, we reviewed the complications associated
with our recent experience of the emergency subtotal
colectomy (ESC).
Patients and methods
Consecutive patients undergoing ESC for colitis at a
district general hospital between 1999 and 2004 were
included. All patients were managed with a closed rectal
(sigmoid) stump, stapled at or above the level of the
pelvic brim. The stump was then either left within the
519
Management of the rectal stump after emergency sub-total colectomy
J. P. Trickett et al.
Table 1 Morbidity, mortality and post operative stay by surgical management of the rectal stump from previous studies.
Surgical
management
of rectal stump
Author [ref]
Carter et al. [7]
Kyle et al. [8]
McKee et al. [9]
Wojdemann et al. [10]
Ng et al. [11]
Mucous fistula
Subcutaneous
Short (intrapelvic)
Combined subcutaneous
and short (intrapelvic)
Intraperitoneal
Intraperitoneal
short (intrapelvic)
Level of pelvic brim
Subcutaneous
Case
number
Pelvic
sepsis rate
(%)
Wound
infection rate
(%)
Stump
discharge rate
(%)
30
55
51
106
7%
4%
12%
7%
0%
13%
12%
35%
23
53
9
147
32
8%
6%
33%
3%
3%
4%
0%
4%
0%
8%
6%
2%
0%
peritoneal cavity or brought up into the subcutaneous
space at the lower end of a midline laparotomy incision,
secured to the fascia of the rectus sheath with interrupted
dissolvable sutures and wound closure completed. The
choice of surgical procedure was determined by individual surgeons’ preference, and did not relate to patients’
pre-operative morbidity or operative findings.
The indications for surgery were acute colitis refractory to medical treatment, toxic megacolon or perforation. Patient data were retrospectively collected on
pre-operative steroid use, inflammatory markers, albumin, indication for surgery, colonic pathology and
surgical management of the rectal stump. Morbidity data
were recorded on rectal stump leak, pelvic sepsis, wound
infection, persistent rectal stump discharge and length of
postoperative hospital stay (POS). Cases were excluded
from POS analysis if complications not related to the
rectal stump extended it. Mann–Whitney and Fischer
exact test statistical analysis was performed using GraphPad InStat (version 3.06).
Results
Thirty-seven ESCs were performed, 34 for disease
refractory to medical treatment, 2 for toxic mega colon
and 1 for perforation. Thirty-four cases were for
ulcerative colitis, 2 Crohn’s colitis and 1 infective colitis.
Mortality
(%)
Post
operative
stay (days)
0%
0%
0%
14
The overall pelvic sepsis rate was 5%. The median POS
was 13 days (5–193 days). Three cases were excluded
from this analysis because of complications not related
to the rectal stump extending POS (intracerebral
haemorrhage, perforated duodenal ulcer and ileostomy
complications).
Twenty-seven patients were managed with an intraperitoneal and 10 a subcutaneously placed closed rectal
stump. There was no pre-operative difference between
the two groups in steroid use or levels of inflammatory
markers. The only statistically significant difference was a
lower median age and a higher albumin level in the
subcutaneous group (Table 2). The only difference in
indications for surgery between the two groups was one
perforation in the intraperitoneal group otherwise in both
groups there was one case of toxic megacolon and the
remainder were cases refractory to medical treatment.
The only exceptions to an open ESC with single layer
stapled closure of the rectal stump were in the intraperitoneal group, 3 cases were performed laparoscopically, 5
had an additional and 3 an exclusively sutured rectal
stump closure (Table 3). The exceptions to a colonic
pathology of ulcerative colitis were 2 cases of Crohn’s
colitis and 1 of infective colitis were also from the
intraperitoneal group. The median POS for patients with
a subcutaneously placed stump was shorter than for those
with an intraperitoneal stump, 8 and 15 days, respectively
Table 2 Pre-operative intergroup standardization by age, steroid treatment, inflammatory markers and albumin.
Steroids treatment
Inflammatory markers
Group
Age (years)
% High dose
Duration (days)
ESR
CRP
WCC
Albumin
Intra peritoneal
Subcutaneous
P-value
43.5
31.5
0.017
68
75
1
13
7
0.08
35
29
0.4
39
30
0.8
9
13
0.1
24.5
32
0.01
520
2005 Blackwell Publishing Ltd. Colorectal Disease, 7, 519–522
Management of the rectal stump after emergency sub-total colectomy
J. P. Trickett et al.
Table 3 Method of stump closure by group.
Stump closure (%)
Group
Stapled
Sewn
Both
Intra peritoneal
Subcutaneous
63
100
14
23
(P ¼ 0.04). Two cases excluded from this analysis were
from the subcutaneous group and one from the intraperitoneal group. Two (7%) patients with UC had leakage
from an intraperitoneal stump, prolonging POS (33 and
193 days); in this group there were 7 (26%) wound
infections and 9 (33%) patients with persistent rectal
discharge. Three (30%) of the subcutaneous stumps
spontaneously opened creating a mucous fistula, in each
case this resulted in a wound infection but did not appear
to prolong POS (6, 7 and 16 days). No patient in this
group had persistent rectal discharge.
Discussion
Creation of a mucous fistula was the original management of the rectal stump after ESC. A number of authors
up to the late 1980s described closure of the rectal stump
after ESC as hazardous because of the risk of pelvic sepsis
[1,3,4]. The only author to report data of higher wound
and pelvic sepsis rates for rectal stump closure attributed
the poorer outcome to the more frequent use of steroid
treatment [5].
There was a reduction in mortality to 3–4% for ESCs
prior to the mid 1980s [1,6], when mucous fistula was
advocated rather than rectal stump closure. However, this
could be attributed to a variety of factors including
improvements in treatment of sepsis rather than just the
use of a second stoma.
Carter et al. [7] in 1991 and Kyle et al. [8] in 1992
reported 7% and 8% pelvic sepsis rates from 106 and 23
cases of rectal stump closure with no mortality. The
wound infection rates were 13% and 4%, respectively, and
in Carter’s study, referred only to the subcutaneous
closures.
McKee et al. [9] reported pelvic sepsis rates of 6% for
62 closed intraperitoneal stumps, with no mortality or
wound infection and a median postoperative stay of
14 days. Wojdenn et al. [10] in 1995 reported a similar
rate of pelvic sepsis from 147 cases of rectal stump closed
at the level of the pelvic brim and an 8% wound infection
rate. The mortality rate of 2% was ascribed to sepsis
related to perforation prior to surgery.
We report a pelvic sepsis rate of 5% for rectal stumps
closed above the level of the recto-sigmoid junction, a
2005 Blackwell Publishing Ltd. Colorectal Disease, 7, 519–522
similar rate to studies reporting no difference in pelvic
sepsis between closed rectal stumps and those brought
out as a mucous fistula [7]. Therefore avoidance of a
second stoma can be achieved without a higher rate of
pelvic sepsis (Table 1).
The incidence of pelvic sepsis when the rectal stump is
closed appears to be related to its location; short
intrapelvic stumps have the highest rates at 33% [7],
intraperitoneally placed stumps 6–12% [7,9] and subcutaneously placed stumps the lowest risk at 3–4% [7,11].
We report a 7% rate for intraperitoneally placed stumps
and no pelvic sepsis associated with subcutaneously
placed stumps. This lower pelvic sepsis rate in subcutaneously placed closed rectal stumps is potentially at the
expense of a higher wound infection rate of 6–13%
[7,11].
A strategy not employed for the study groups but with
the potential to reduce post ESC sepsis is routine
postoperative anal drainage of the rectal stump. A lower
rate of stump discharge and pelvic sepsis has been
suggested [11] for routinely drained subcutaneously
placed closed rectal stumps.
Greater morbidity, reflected in prolonged in-patient
stay, would be expected following intraperitoneal leakage
and pelvic sepsis rather than subcutaneous stump leakage
or wound infection.
The only previous study to compare subcutaneously
with intraperitoneal closed stumps [7] reported a lower
pelvic sepsis rate of 4% vs 12% but a higher wound
infection rate of 13%, with more than a third having
breakdown of the stump closure in the subcutaneously
placed group. The two groups were managed differently,
ESC and subcutaneously placed stumps were performed
at a specialist tertiary referral centre, the intraperitoneally
placed group however, were managed at a variety of
district hospitals. No assessment of overall morbidity in
terms of POS was made.
We have observed a lower rate of pelvic septic
complications and shorter in hospital stay for subcutaneously vs intraperitoneally placed closed rectal stumps,
despite extraperitoneal stump discharge and a higher
wound infection rate, thus suggesting an overall lower
morbidity rate for subcutaneously placed closed rectal
stumps.
The infrequency of ESCs for severe colitis, means
single centre studies yield too few patients to provide
robust evidence of difference between techniques.
Rectal stump closure is acceptable practice in ESC for
severe colitis. Subcutaneous placement of a closed stump
may reduce total morbidity.
A randomised prospective multicentre trial incorporating postoperative anal drainage of the rectal stump is
required to confirm significance.
521
Management of the rectal stump after emergency sub-total colectomy
References
1 Hawley PR. Emergency surgery for ulcerative colitis. World J
Surg 1988; 12: 169–73.
2 Glotzer DJ. Operations for inflammatory bowel disease:
indications and type. Clin Gastroenterol 1980; 9: 371–87.
3 Fazio VW, Turnbull RB. Ulcerative colitis and Crohn’s
disease of the colon. Med Clin North Am 1980; 64: 1134–
59.
4 Lee ECG, Truelove SC. Proctocolectomy for ulcerative
colitis. World J Surg 1980; 4: 195–201.
5 Koudahl G, Aagaard P. The management of the rectal stump
after subtotal colectomy for ulcerative colitis. Scand J
Gastroenterol 1971; 9: 127–9.
6 Leijonmarck C-E, Brostrom O, Monsen U, Hellers G.
Surgical treatment of ulcerative colitis in Stockholm county,
1955–84. Dis Colon Rectum 1989; 32: 918–26.
522
J. P. Trickett et al.
7 Carter FM, McLeod RS, Cohen Z. Subtotal colectomy for
ulcerative colitis: complications related to the rectal remnant.
Dis Colon Rectum 1991; 34: 1005–9.
8 Kyle SM, Steyn RS, Keenan RA. Management of the rectum
following colectomy for acute colitis. Aust NZ J Surg 1992;
62: 196–9.
9 McKee RF, Keenan RA, Munro A. Colectomy for acute
colitis: is it safe to close the rectal stump? Int J Colorect Dis
1995; 10: 222–4.
10 Wojdemann M, Wettergren A, Hartvigsen A, Myrhol T,
Svendsen LB, Bulow S. Closure of rectal stump after
colectomy for acute colitis. Int J Colorect Dis 1995; 10:
197–9.
11 Ng RLH, Davies AH, Grace RH, Mortensen NJ, McC.
Subcutaneous rectal stump closure after emergency subtotal
colectomy. Br J Surg 1992; 79: 701–3.
2005 Blackwell Publishing Ltd. Colorectal Disease, 7, 519–522