Loop Versus Divided Colostomy for the Management of Anorectal Malforma-
tions: A Systematic Review And Meta-analysis

Fouad Youssef, Ghaidaa Arbash, Pramod S. Puligandla, Robert J Baird

PII: S0022-3468(17)30077-5
DOI: doi: 10.1016/j.jpedsurg.2017.01.044
Reference: YJPSU 58020

To appear in: Journal of Pediatric Surgery

Received date: 17 January 2017

Accepted date: 23 January 2017

Please cite this article as: Youssef Fouad, Arbash Ghaidaa, Puligandla Pramod S., Baird
Robert J, Loop Versus Divided Colostomy for the Management of Anorectal Malforma-
tions: A Systematic Review And Meta-analysis, Journal of Pediatric Surgery (2017), doi:
10.1016/j.jpedsurg.2017.01.044

This is a PDF le 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 nal form. Please note that during the production process
errors may be discovered which could aect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Loop Versus Divided Colostomy for the Management of Anorectal Malformations: A

Systematic Review And Meta-analysis

P T
Fouad Youssef MD, Ghaidaa Arbash MD, Pramod S. Puligandla MD MSc,

RI
Robert J Baird MDCM, MSc*

SC
The Division of Pediatric General and Thoracic Surgery, The Montreal Childrens Hospital of the McGill
University Health Centre, Montreal, Quebec, Canada, H4A 3J1
*Corresponding Author:

NU
Robert J. Baird, MDCM, MSc
Division of Pediatric General and Thoracic Surgery
The Montreal Childrens Hospital
MA
1001 Decarie Boulevard, Montreal, QC H4A 3J1
Email: robert.baird@mcgill.ca
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Abstract

T
Background

P
The ideal colostomy type for patients with anorectal malformations (ARM) is undetermined. We

RI
performed a systematic review and meta-analysis of short-term complications comparing loop
and divided colostomies.

SC
Methods
After review registration (PROSPERO: CRD42016036481), multiple databases were searched

NU
for comparative studies without language or date restrictions. Grey literature was sought.
Complications investigated included stomal prolapse/hernia/retraction, wound infections, and
MA
urinary tract infections (UTIs). Two reviewers independently assessed study eligibility and the
quality of included studies. Meta-analysis of selected complications was performed using
ED

Revman 5.3, with p<0.05 considered significant.

Results
PT

Twenty-six studies were included, and four were multi-institutional. Reporting standards were
highly variable. Studies scored between 6-9 of possible nine stars on the NOS. Overall, 3866
CE

neonates with ARM were incorporated, in which 2241 loop colostomies and 1994 divided
colostomies were reported. Of 10 studies reporting short-term complications, the overall rate was
AC

27%. Meta-analysis demonstrated no significant difference in the incidence of UTIs, (OR: 2.55
[0.76, 8.58], p=0.12), while loop colostomies had a significantly higher prolapse rate (See
figure). No publication bias was noted.
Conclusions
A colostomy for patients with an ARM is a source of considerable morbidity. Divided
colostomies reduce the risk of subsequent prolapse and may represent the preferred approach.

Keywords: Anorectal malformation, loop, divided, systematic review, meta-analysis.

Level of Evidence: 3A
 ACCEPTED MANUSCRIPT

Introduction

Anorectal malformations (ARM) encompass a wide range of congenital deformities that

T
often necessitate complex reconstructive skills. Prior to definitive repair, colostomies are

P
commonly performed to divert the fecal stream. [1] Two variations of colostomy are performed;

RI
loop, and divided colostomies, both of which have considerable risks of post-operative
complications related to the stoma site (prolapse, retraction, stenosis) or inadequacy of diversion

SC
(urinary infections). [2] A loop colostomy involves fixing an opened intestinal loop (typically
descending colon) to one incision in the abdominal wall without complete transection of the

NU
intestine. On the other hand, a divided colostomy entails complete division of the bowel loop
resulting in a proximal colostomy and separated distal mucous fistula that are typically sutured to
MA
the abdominal wall through separate incisions. While the traditional approach has favored a
divided colostomy, [3] newer publications seem to suggest comparable results using a loop. [4]
ED

Both stoma options are associated with specific merits and drawbacks and no consensus
has been reached regarding the optimal procedure for a given anatomic configuration. The
PT

relative ease of creation and eventual takedown of the loop colostomy is generally balanced
against the more definitive diversion associated with a true divided colostomy. As such, the aim
CE

of this study is to compare the short-term outcomes between the loop and divided colostomies in
patients with anorectal malformation via a formal systematic review and meta-analysis.
AC

1. Materials and Methods

1.1. Search Strategy

The protocol of this review was prospectively registered in PROSPERO (CRD42016036481),

(http://www.crd.york.ac.uk/PROSPERO/). With no date or language restrictions, the following
online sources were searched: AMED, PubMed, Africa-wide information, EMBASE, Global
Health, Biosys, MEDLINE, Cochrane, LLAOS, and Web of Science in January 2016. With the
guidance of our institutions scientific librarian, each of these databases was approached through
a particular search strategy in accordance with the database design (Appendix). The reference
lists of the included studies were also hand-checked for any pertinent articles. Grey literature was
 ACCEPTED MANUSCRIPT

sought, by exploring the abstracts of the relevant conferences and contacting available authors
for unpublished data when possible. Only comparative studies were included in the qualitative
analyses, and of these comparative studies, only studies that looked solely at ARM were included
in the quantitative analyses.

P T
1.2. Definitions

RI
Through the course of this review, the following definitions were used:

SC
A. Loop Colostomy: The use of the opened bowel loop as a stoma without the complete splitting
of the intestinal wall which resulted in one stoma opening on the skin.

NU
B. Divided Colostomy: The complete splitting of the bowel loop resulting in two stoma openings
on the skin with two separated skin incisions.
MA
1.3. Inclusion and Exclusion criteria

According to PICOS format, inclusion criteria for this review were the following:
ED

Population: All infants born alive with the diagnosis of ARM and who underwent either
PT

loop or divided colostomy at any point during their management.

Intervention: Loop colostomy irrespective of the portion of large bowel where the stoma
CE

was performed.
Control: Divided split colostomy irrespective of the portion of large bowel where the
AC

stoma was performed.

Outcomes: The primary outcome was the incidence of stoma prolapse. Secondary
outcomes included the following stoma-related complications: urinary tract infection,
para-stomal hernia, skin excoriation, stoma retraction, wound infection, stoma necrosis,
stoma stricture, and stoma bleeding.
Study type: Only comparative studies that incorporated newborns with ARM were
included in this review; studies that looked only at ARM patients were included for
quantitative meta-analyses.

Two investigators twice examined the resulted titles independently; the following exclusion
criteria were applied: any study that did not include ARM patients, basic science or animal
reports, case reports, and non-comparative studies. Identified abstracts were examined and
 ACCEPTED MANUSCRIPT

further exclusions were applied to all non-relevant reports and to all the studies that did not
contain either the primary or the secondary outcomes of this review. Full texts were retrieved
for the studies that were considered pertinent by either reviewer. Full consensus of
incorporated studies was then reached through discussion under the guidance of the senior

T
author (RB).

P
RI
1.3. Quality of included studies

SC
Two reviewers independently evaluated the quality of included studies. The Critical Appraisal
Skills Programme (CASP) Randomised Controlled Trials Checklist was used to evaluate available

NU
randomized controlled trials (RCT); [5] The Newcastle-Ottawa Scale (NOS) for case-control
studies was used to critically appraise the quality of included case-control papers. [6] The NOS is
designed to assess study quality in three broad domains: the selection of the study groups,
MA
comparability between the groups, and ascertainment of exposure. Stars are given for each of the
quality items met by the study; every study gains a score of a maximum nine stars.
ED

1.4. Statistical analysis

PT

Meta-analysis was performed for primary and selected secondary outcomes using
CE

RevMan 5.3 (Copenhagen). The random effect model (Mantel-Haenszel approach) was used to
create standard forest plots of effect size and error bars, with heterogeneity reported for each
AC

analysis. Publication bias was evaluated through the generation of a funnel plot of standard error
against the log odds ratio, although a statistical evaluation of this plot was not undertaken.
P<0.05 was considered statistically significant for all analyses.

2. Results

The search results are shown in the PRISMA flow chart in Figure 1. Out of 3129 titles found
by online search and 8 by hand search, 1777 records were identified after duplicate removal.
Title and abstract screening subsequently excluded 1737 records. Forty full-text papers were
retrieved for further evaluation, of which 14 were also excluded. Finally, twenty-six studies were
included. Two prospective [7,8], one randomised control study, [4] and 23 retrospective studies
 ACCEPTED MANUSCRIPT

[1,2,9-29] were included for qualitative analysis, of which, 13 were included for quantitative
analysis.

2.1 Qualitative analysis

T
The characteristics of the included studies are highlighted in Table 1. The years of publication

P
spanned the era from 1980 to 2016. Sample sizes varied between 9 and 1470 cases of

RI
ARM/study. Four studies were multi-institutional. [1,10,15,23] Overall, 14 studies included
patients with indications for colostomy other than the ARM (like Hirschsprungs disease) while

SC
the other 12 studies looked at the complications of colostomy only in ARM patients. A total of

NU
3866 neonates with ARM were incorporated; in which 2241 loop colostomies and 1994 divided
colostomies were reported. The overall average complication rate through the 26 studies was
27%. Out of the 26 included studies in this review, only 13 studies provided high quality data
MA
comparing loop and divided colostomy in ARM patients [1,4,7,11,13,15,17,20,22,23,25,28,29].
The other 13 studies included mixed indications of colostomy in addition to ARM rendering
ED

quantitative analysis impossible.

The quality of included studies is shown in Table 2. Included studies were of moderate to high
PT

quality, scoring between 6 and 9 out of possible nine stars.

CE

2.2 Quantitative analysis

2.2.1. Stoma Prolapse

AC

Nine studies reported the incidence of prolapse as a complication of colostomy in patients with
ARM. [1,7,13,17,20,22,23,25,29] The data from all but one of these 9 studies [7] was
appropriate to be included for the meta-analysis. Three studies favoured the divided colostomy
(less prolapse) [1,20,23] while the other five demonstrated comparable results for prolapse
between the two groups. Pooled data (2137 patients) favored the divided colostomy over the loop
colostomy for the incidence of stoma prolapse. (OR: 2,34; 95% CI: 1.01 to 5.47; P=0.05)
(Figure2A). Heterogeneity between studies was found to be significant (I2=67%). No definition
of stoma prolapse was noted in any of the included studies. A Funnel plot of the log odds ratio
against the standard error for these eight studies is demonstrated in Figure 2b; its relative
asymmetry suggests possible publication bias.
 ACCEPTED MANUSCRIPT

T
2.2.2. Urinary tract infection

P
RI
Five studies evaluated the incidence of Urinary Tract Infection (UTI) after stoma creation
[17,20,22,23,25] of which only one study favoured divided stoma. [23] The other four studies

SC
demonstrated comparable rates of UTI between the two groups. Of note, the authors of the most
recently published study in this review [17] performed a risk-stratification of the ARM patients

NU
based on the presence or absence of a recto-urinary fistula. Despite this, they found that the risk
of UTI was still independent of the stoma type. Only two studies [17,20] included a definition of
MA
UTI which was defined as a combination of suggestive clinical symptoms and a positive urine
culture. Data from all the five studies were used for meta-analysis of the pooled cohort (977
patients). Meta-analysis showed no significant difference in the incidence of UTI between the
ED

two groups with relatively significant heterogeneity (I2=85%) (Figure 3).

PT

2.2.3. Skin excoriation

Three studies had data regarding skin excoriation, [17,25,28]. Only one study individually
CE

favoured loop stoma [28] while the other two showed comparable results. No definition of skin
excoriation was presented in any of these three studies and meta-analysis of 340 included
AC

patients revealed no difference in the incidence of skin excoriation between the two groups
(Figure 4).

2.2.4. Stoma retraction

Three studies compared the incidence of stoma retraction between loop and divided cohorts.
[17,20,25]. All three studies showed comparable results between the two groups. None of these
three studies defined stoma retraction. With 386 included patients, the meta-analysis showed no
difference in the incidence of stoma retraction between the two groups. Considerable
heterogeneity was noted across the included studies (I2 = 60%) (Figure 5).

2.2.5 Para-stomal hernia

 ACCEPTED MANUSCRIPT

Four studies reported the incidence rates of para-stomal hernia [13,17,20,25] and no significant
difference was reported between the two groups in any of these four studies. Furthermore, none
of these four studies adequately defined para-stomal hernia. With a total of 523 included patients,
the meta-analysis revealed comparable results between the two groups (Figure 6).

P T
RI
2.2.6 Wound infection

SC
Three studies reported the incidence rates of post-intervention wound infection. [17,25,29] of
which, one study [17] defined the wound infection as a cellulitis that required treatment either

NU
with antibiotic or opening of the incision. With a total of 394 patients, meta-analysis
demonstrated that the incidence of post-intervention wound infection was comparable between
the two groups (Figure 7).
MA
2.2.7. Stoma stricture

Three studies showed data on the incidence rates of stoma stricture, [17,25,29] none of which
ED

contained a definition of stoma stricture. Meta-analysis of pooled 394 patients showed no

significant difference between the two groups. (Figure 8).
PT

2.2.8 Stoma necrosis

CE

Two studies provided data on the incidence of stoma necrosis. [17,25] Definitions of stoma
necrosis were not provided in either study. Both of these studies showed no difference in the
AC

incidence of stoma necrosis between groups.

2.2.9 Other complications

None of the included studies reported high quality data that allowed comparisons of loop and
divided colostomies regarding the incidence of the following complications: mortality, stoma
bleeding, stoma obstruction and stoma mislocation. The studies included in this review did not
address results of the definitive ARM repair.

Discussion:
 ACCEPTED MANUSCRIPT

The creation of a stoma is often the first stage of a complex surgical treatment plan for a patient
with an anorectal malformation (ARM). While the quality of the ultimate surgical reconstruction
results in ramifications that last a lifetime, the short period of time with a stoma is nonetheless a
critical period in the early treatment of ARM. Complications resulting from improper stoma

T
creation can lead to additional (and unnecessary) urgent procedures, and even life-threatening

P
infections. The decision regarding the type of stoma should ideally be informed by the best

RI
available evidence that reduces these adverse events. In 2014, Van den Hondel et al. reviewed

SC
the complication rates of stomas for patients with ARMs and Hirschsprungs disease. [29]
Although this report listed the most common colostomy-related complications, the pooled data

NU
lacked a formal meta-analysis and failed to distinguish between colostomies based on indication.
Thus, a systematic review and meta-analysis evaluating data specific to ARM is needed.
MA
Furthermore, several more recent studies continue to challenge existing dogma surrounding
colostomies for ARMs. [17,20]. As such, a more disease-specific and exhaustive review was
undertaken.
ED

The traditional approach to a newborn child with an ARM requiring immediate fecal diversion
PT

has been a divided colostomy as championed by Pena et al. As originally described, this involves
a generous left lower quadrant incision and laparotomy, followed by intestinal division and
CE

fixation at either end of the incision after fascial closure. [3] There are several advantages to this
approach, including the ability to evaluate and treat concomitant pathology like hydrocolpos,
AC

ensuring proper orientation of the mesentery and stoma, as well as the elimination of
downstream spillage by placing the stoma appliance away from the mucous fistula. The chief
disadvantage is the relative invasiveness of the initial procedure, the consequent potential for
intra-abdominal adhesion formation, and the requirement of a laparotomy for stoma takedown.
By comparison, a loop colostomy can be performed through a small incision and requires a very
limited procedure for eventual takedown. However, spillage of fecal contents into the distal limb
may result in urinary infections or megarectosigmoid. There is also the potential for
malorientation of the stomas or mesentery through a small incision, although this may be
obviated via laparoscopic visualization. [30]. Irrespective of the chosen technique, great care
must be taken to ensure proper positioning of the stoma (avoiding transverse colostomies if
possible), a sufficiently long distal rectosigmoid segment to allow for tension free reconstruction,
 ACCEPTED MANUSCRIPT

proper size of the fascial aperture and adequate fixation of the stoma to the abdominal wall.
However, the absolute risk of these and other complications is unclear. As a result, the choice of
stoma type has largely remained individual or institution specific.

T
Our analysis clearly demonstrates that neonatal colostomies created for the treatment of anorectal

P
malformations are associated with considerable risk of postoperative adverse events.

RI
Furthermore, contrary to conventional wisdom, urinary infections were not significantly higher
in patients receiving a loop colostomy when compared to a divided colostomy. It is possible that

SC
an insufficient number of patients were analyzed to demonstrate a difference, although our large
number of included patients (approximately 500 patients in each arm) suggests that concern for

NU
spillover into the distal limb may be unfounded. Many individuals place a rod spur underneath
the intestinal loop, which may sufficiently kink the loop as to minimize spillover. An additional
MA
technical modification involves raising the proximal stoma over the surface of the skin
(Brooking) and narrowing down the distal lumen in order to further reduce the potential for
ED

spillover of colonic contents. Although this review could not evaluate these and other technical
details of stoma creation, fear of urinary infections alone may not be a justified reason to proceed
PT

with a divided colostomy.

On the other hand, our review did demonstrate a significant difference in the rate of prolapse
CE

favoring divided colostomies, with nearly 1000 patients in each arm. This is not altogether
surprising given that the larger abdominal incision more easily affords the opportunity to
AC

adequately secure the bowel to the fascia and likely results in more intra-abdominal scarring. In
addition, it is possible that colonic peristalsis may more readily disrupt points of fixation when
the intestine is configured as a loop as opposed to an end colostomy. The burden of stomal
prolapse is challenging to quantify, however, as included studies failed to indicate whether its
management required simple manual reduction or urgent operative repair. Our review revealed
no other significant differences in shortterm complication rates, including retraction, necrosis,
parastomal hernia or peri-stomal skin breakdown.

In considering the clinical significance of an increased prolapse rate with loop colostomies, it is
important to weigh the morbidity of prolapse with the advantages of a loop colostomy. Colonic
prolapse certainly does not necessarily mandate an unplanned operation, with many simply being
 ACCEPTED MANUSCRIPT

managed with periodic bedside reduction or even simple observation until the time of stoma
closure. However, intestinal prolapse is typically distressing both to caregivers and health
professionals delivering care. It may bleed as a result of congestion and/or mucosal irritation if it
rubs against the appliance, thus occasionally requiring stoma revision. Our analysis failed to

T
reveal the percentage of stomal prolapse requiring intervention due to the lack of consistent

P
reporting. Furthermore, no identified report provided a comparison of resource utilization related

RI
to the type of stoma either via an analysis of direct costs per procedure or the surrogate of

SC
operating room time.

Several additional limitations of this review require acknowledgment and render judgment of

NU
best practice unclear. The quality of this systematic review is dependent on the collection of
studies contained within. With only one small RCT and extreme differences in reporting
MA
practices across all included studies, firm conclusions cannot be drawn with the available dataset.
Publication bias may also be present, although efforts to access grey literature were undertaken.
ED

In addition, this review encompassed an extremely wide temporal and geographic variety of
studies, aggregating publications from different eras and very different societal contexts. While
PT

ARMs have always presented in a similar fashion, differences in care practices over time may
have diluted our reported outcomes thus preventing us from identifying differences when clinical
CE

variation actually existed. Relatively high statistical heterogeneity between studies supports this
possibility. In addition, the analysis of short-term complications related to stoma type fails to
AC

address the differences that arise during stoma takedown or thereafter. These could include the
rate of immediate complications after surgery, differences in operating room times and also the
rate of adhesive bowel obstruction in the long-term.

Available evidence concerning the significant short-term complication rate of colostomy creation
for the patient with ARMs suggests that a detailed informed consent process is required and that
the surgeon must aspire to technical excellence regardless of preferred stoma type. While loop
colostomies can be performed and reversed in a more expeditious manner, they are associated
with an increased risk of prolapse. In the absence of a clearly superior choice, clinicians must
choose the stoma type based on their personal comfort with procedure, as well as a detailed
description with the childs caregiver concerning the advantages and drawbacks inherent with
each option. Further research is required to illuminate whether patient-related factors such as
 ACCEPTED MANUSCRIPT

malformation sub-type, age or stoma location alters complication rates and should dictate choice
of stoma.

T
Acknowledgments:

P
RI
We are indebted to Elena Guadagno, the librarian in the Montreal Childrens Hospital for her
help with the literature search.

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

References:
[1] Wilkins S, Pea A: The role of colostomy in the management of anorectal malformations.
Pediatr Surg Int1988;3:105-109.
[2] Mollitt DL, Malangoni MA, Ballantine TV, et al: Colostomy complications in children. An

T
analysis of 146 cases. Arch Surg 1980;115:455-458.

P
[3] Levitt M, Pena A. Anorectal Malformations. In Coarn A, Adzick NS, Krummel TM et al.

RI
Pediatric Surgery 7th ed., Philadelphia: Elsevier Saunders; 2012 p. 1289-1310.

SC
[4] Askarpour S, Peyvasteh M, Changai B, et al: Skin bridge versus rod colostomy in children -
comparison between complications. Pol Przegl Chir 2012;84:485-487.

NU
[5] Critical Appraisal Skills Programme (CASP) www.casp.uk.net [accessed 16.05.30].
[6] Wells GA, Shea B, O'Connell D, et al. (2011) The Newcastle-Ottawa Scale (NOS) for
MA
assessing the quality of nonrandomized studies in meta-analysis.
www.ohri.ca/programs/clinical_epidemiology/oxford.asp [accessed 16.05.30]
[7] Khan K, Younas M, Waheed T: Management of colostomies in infancy. Journal of
ED

Postgraduate Medical Institute (Peshawar-Pakistan) 2011;17.

[8] Dode LGC: Childhood colostomy and its complications in Lagos. 2007.
PT

[9] Chandramouli B, Srinivasan K, Jagdish S, et al: Morbidity and mortality of colostomy and its
closure in children. J Pediatr Surg 2004;39:596-599.
CE

[10] Chirdan LB, Uba FA, Ameh EA, et al: Colostomy for high anorectal malformation: an
evaluation of morbidity and mortality in a developing country. Pediatr Surg Int 2008;24:407-410.
AC

[11] Chowdhary SK, Chalapathi G, Narasimhan KL, et al: An audit of neonatal colostomy for
high anorectal malformation: the developing world perspective. Pediatr Surg Int 2004;20:111-
113.
[12] Cigdem MK, Onen A, Duran H, et al: The mechanical complications of colostomy in infants
and children: analysis of 473 cases of a single center. Pediatr Surg Int 2006;22:671-676.
[13] Demirogullari B, Yilmaz Y, Yildiz GE, et al: Ostomy complications in patients with
anorectal malformations. Pediatr Surg Int 2011;27:1075-1078.
[14] Ekenze SO, Agugua-Obianyo NE, Amah CC: Colostomy for large bowel anomalies in
children: a case controlled study. Int J Surg 2007;5:273-277.
 ACCEPTED MANUSCRIPT

[15] Figueroa M, Bailez M, Solana J: [Colostomy morbidity in children with anorectal

malformations (ARM)]. Cirugia pediatrica: organo oficial de la Sociedad Espanola de Cirugia
Pediatrica 2007;20:79-82.
[16] Gardikis S, Antypas S, Mamoulakis C, et al: Colostomy type in anorectal malformations:

T
10-years experience. Minerva Pediatr 2004;56:425-429.

P
[17] Liechty ST, Barnhart DC, Huber JT, et al: The morbidity of a divided stoma compared to a

RI
loop colostomy in patients with anorectal malformation. J Pediatr Surg 2016;51:107-110.

SC
[18] Mohammad MA, Al Amin O: A Look into Colostomy in Pediatric Patients Presented To Al
Ribat University Hospital December 2013-December 2014. Sudan Journal of Medical Sciences

NU
2015;10:21-24.
[19] Nour S, Beck J, Stringer MD: Colostomy complications in infants and children. Ann R Coll
MA
Surg Engl 1996;78:526-530.
[20] Oda O, Davies D, Colapinto K, et al: Loop versus divided colostomy for the management of
anorectal malformations. J Pediatr Surg 2014;49:87-90.
ED

[21] Osifo O, Osaigbovo E, Obeta E: Colostomy in children: Indications and common problems
in Benin City, Nigeria. PAKISTAN JOURNAL OF MEDICAL SCIENCES 2008;24:199.
PT

[22] Patwardhan N, Kiely EM, Drake DP, et al: Colostomy for anorectal anomalies: high
incidence of complications. J Pediatr Surg 2001;36:795-798.
CE

[23] Pena A, Migotto-Krieger M, Levitt MA: Colostomy in anorectal malformations: a procedure

with serious but preventable complications. J Pediatr Surg 41:748-756; discussion 2006;748-756.
AC

[24] Rees BI, Thomas DF, Negam M: Colostomies in infancy and childhood. Z Kinderchir
1982;36:100-102.
[25] Sheikh MA, Akhtar J, Ahmed S: Complications / problems of colostomy in infants and
children. J Coll Physicians Surg Pak 2006;16:509-513.
[26] Soomro BA, Solangi RA, Siddiqui MA: Colostomy in children: Indications and
complications. Pak J Med Sci October-December 2010;26:883-886.
[27] Uba A, Chirdan L: Colostomy complications in children. Annals of African medicine
2003;2:9-12.
[28] AlAssiri A, Almosallam O, AlShanafey S, et al: Outcome of loop versus divided colostomy
in the management of anorectal malformations.
 ACCEPTED MANUSCRIPT

[29] Van den Hondel D, Sloots C, Meeussen C, et al: To split or not to split: colostomy
complications for anorectal malformations or hirschsprung disease: a single center experience
and a systematic review of the literature. Eur J Pediatr Surg 2014;24:61-69.
[30] Liem NT, Quynh TA. Single trocar laparoscopic-assisted colostomy in newborns. Pediatr

T
Surg Intl 2013;29:651-3

P
RI
SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Legends for illustrations:

Figure 1: PRISMA Flow Diagram From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group

(2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS

P T
Med 6(6): e1000097. doi:10.1371/journal.pmed1000097 For more information, visit www.prisma-

RI
statement.org.

SC
Table 1. Characteristics of included studies; Ret. = Retrospective, SC= Single center, MC= Multi-center,

ARM= Anorectal malformation, UTI= urinary tract infection, Pro= Prospective, Ret= Retrospective.

NU
Table 2. The Newcastle Ottawa Scale (NOS) for assessing the quality of included studies.
MA
Figure 2. A: Forest plot of prolapse rate comparing loop and divided colostomy groups.
ED

Figure 2. B: Funnel plot of prolapse rate.

PT

Figure 3. Forest plot of urinary tract infection comparing loop and divided groups.

Figure 4. Forest plot of skin excoriation comparing loop and divided groups.
CE

Figure 5. Forest plot of stoma retraction comparing loop and divided groups.
AC

Figure 6. Forest plot of parastomal hernia comparing loop and divided groups.

Figure 7. Forest plot of wound infection comparing loop and divided groups.

Figure 8. Forest plot of stoma stricture comparing loop and divided groups.
 ACCEPTED MANUSCRIPT

Appendix: Search strategy (Pubmed).

PubMed [NLM] (January 12, 2016)

P T
Search Query Items
#23 #20 NOT #19 7

RI
#22 #19 NOT #20 11
#21 #19 AND #20 3

SC
#20 #16 AND #17 10
#19 #15 AND #17 14

NU
#18 #14 AND #17 21
#17 ((inprocess[sb]0 OR (publisher[sb] NOT pmcbook))) 479574
#16 #11 NOT #13 MA 802
#15 #5 NOT #13 1064
#14 #12 NOT #13 1700
(animals[Title] OR animal[Title] OR mice[Title] OR mus[Title] OR mouse[Title] OR murine[Title] OR woodmouse[Title] OR rats[Title] OR rat[Title] OR murinae[Title] OR muridae[Title] OR cottonrat[Title] OR cottonrats[Title]
OR hamster[Title] OR hamsters[Title] OR cricetinae[Title] OR rodentia[Title] OR rodent[Title] OR rodents[Title] OR pigs[Title] OR pig[Title] OR porcine[Title] OR swine[Title] OR swines[Title] OR piglets[Title] OR
piglet[Title] OR boar[Title] OR boars[Title] OR "sus scrofa"[Title] OR ferrets[Title] OR ferret[Title] OR polecat[Title] OR polecats[Title] OR "mustela putorius"[Title] OR "guinea pigs"[Title] OR "guinea pig"[Title] OR
cavia[Title] OR callithrix[Title] OR marmoset[Title] OR marmosets[Title] OR cebuella[Title] OR hapale[Title] OR octodon[Title] OR chinchilla[Title] OR chinchillas[Title] OR gerbillinae[Title] OR gerbil[Title] OR gerbils[Title]
OR jird[Title] OR jirds[Title] OR merione[Title] OR meriones[Title] OR rabbits[Title] OR rabbit[Title] OR hares[Title] OR hare[Title] OR diptera[Title] OR flies[Title] OR fly[Title] OR dipteral[Title] OR drosphila[Title] OR
drosophilidae[Title] OR cats[Title] OR cat[Title] OR carus[Title] OR felis[Title] OR nematoda[Title] OR nematode[Title] OR nematoda[Title] OR nematode[Title] OR nematodes[Title] OR sipunculida[Title] OR dogs[Title] OR
ED

dog[Title] OR canine[Title] OR canines[Title] OR canis[Title] OR sheep[Title] OR sheeps[Title] OR lamb[Title] OR lambs[Title] OR mouflon[Title] OR mouflons[Title] OR ovis[Title] OR goats[Title] OR goat[Title] OR
capra[Title] OR capras[Title] OR rupicapra[Title] OR chamois[Title] OR haplorhini[Title] OR monkey[Title] OR monkeys[Title] OR macaque[Title] OR macaques[Title] OR primate[Title] OR primates[Title] OR
anthropoidea[Title] OR anthropoids[Title] OR saguinus[Title] OR tamarin[Title] OR tamarins[Title] OR leontopithecus[Title] OR hominidae[Title] OR ape[Title] OR apes[Title] OR pan[Title] OR paniscus[Title] OR "pan
paniscus"[Title] OR bonobo[Title] OR bonobos[Title] OR troglodytes[Title] OR "pan troglodytes"[Title] OR gibbon[Title] OR gibbons[Title] OR siamang[Title] OR siamangs[Title] OR nomascus[Title] OR symphalangus[Title]
OR chimpanzee[Title] OR chimpanzees[Title] OR prosimians[Title] OR "bush baby"[Title] OR prosimian[Title] OR "bush babies"[Title] OR galagos[Title] OR galago[Title] OR pongidae[Title] OR gorilla[Title] OR
gorillas[Title] OR pongo[Title] OR pygmaeus[Title] OR "pongo pygmaeus"[Title] OR orangutans[Title] OR pygmaeus[Title] OR lemur[Title] OR lemurs[Title] OR lemuridae[Title] OR horse[Title] OR horses[Title] OR
pongo[Title] OR equus[Title] OR cow[Title] OR cows[Title] OR calf[Title] OR bull[Title] OR chicken[Title] OR chickens[Title] OR gallus[Title] OR quail[Title] OR bird[Title] OR birds[Title] OR quails[Title] OR poultry[Title] OR
poultries[Title] OR fowl[Title] OR fowls[Title] OR reptile[Title] OR reptilia[Title] OR reptiles[Title] OR snakes[Title] OR snake[Title] OR lizard[Title] OR lizards[Title] OR alligator[Title] OR alligators[Title] OR crocodile[Title]
PT

OR crocodiles[Title] OR turtle[Title] OR turtles[Title] OR amphibian[Title] OR amphibians[Title] OR amphibia[Title] OR frog[Title] OR frogs[Title] OR bombina[Title] OR salientia[Title] OR toad[Title] OR toads[Title] OR
"epidalea calamita"[Title] OR salamander[Title] OR salamanders[Title] OR eel[Title] OR eels[Title] OR fish[Title] OR fishes[Title] OR pisces[Title] OR catfish[Title] OR catfishes[Title] OR siluriformes[Title] OR arius[Title]
OR heteropneustes[Title] OR sheatfish[Title] OR perch[Title] OR perches[Title] OR percidae[Title] OR perca[Title] OR trout[Title] OR trouts[Title] OR char[Title] OR chars[Title] OR salvelinus[Title] OR "fathead
minnow"[Title] OR minnow[Title] OR cyprinidae[Title] OR carps[Title] OR carp[Title] OR zebrafish[Title] OR zebrafishes[Title] OR goldfish[Title] OR goldfishes[Title] OR guppy[Title] OR guppies[Title] OR chub[Title] OR
chubs[Title] OR tinca[Title] OR barbels[Title] OR barbus[Title] OR pimephales[Title] OR promelas[Title] OR "poecilia reticulata"[Title] OR mullet[Title] OR mullets[Title] OR seahorse[Title] OR seahorses[Title] OR "mugil
curema"[Title] OR "atlantic cod"[Title] OR shark[Title] OR sharks[Title] OR catshark[Title] OR anguilla[Title] OR salmonid[Title] OR salmonids[Title] OR whitefish[Title] OR whitefishes[Title] OR salmon[Title] OR
salmons[Title] OR sole[Title] OR solea[Title] OR "sea lamprey"[Title] OR lamprey[Title] OR lampreys[Title] OR pumpkinseed[Title] OR sunfish[Title] OR sunfishes[Title] OR tilapia[Title] OR tilapias[Title] OR turbot[Title] OR
CE

turbots[Title] OR flatfish[Title] OR flatfishes[Title] OR sciuridae[Title] OR squirrel[Title] OR squirrels[Title] OR chipmunk[Title] OR chipmunks[Title] OR suslik[Title] OR susliks[Title] OR vole[Title] OR voles[Title] OR
lemming[Title] OR lemmings[Title] OR muskrat[Title] OR muskrats[Title] OR lemmus[Title] OR otter[Title] OR otters[Title] OR marten[Title] OR martens[Title] OR martes[Title] OR weasel[Title] OR badger[Title] OR
badgers[Title] OR ermine[Title] OR mink[Title] OR minks[Title] OR sable[Title] OR sables[Title] OR gulo[Title] OR gulos[Title] OR wolverine[Title] OR wolverines[Title] OR minks[Title] OR mustela[Title] OR llama[Title] OR
llamas[Title] OR alpaca[Title] OR alpacas[Title] OR camelid[Title] OR camelids[Title] OR guanaco[Title] OR guanacos[Title] OR chiroptera[Title] OR chiropteras[Title] OR bat[Title] OR bats[Title] OR fox[Title] OR
foxes[Title] OR iguana[Title] OR iguanas[Title] OR "xenopus laevis"[Title] OR parakeet[Title] OR parakeets[Title] OR parrot[Title] OR parrots[Title] OR donkey[Title] OR donkeys[Title] OR mule[Title] OR mules[Title] OR
zebra[Title] OR zebras[Title] OR shrew[Title] OR shrews[Title] OR bison[Title] OR bisons[Title] OR buffalo[Title] OR buffaloes[Title] OR deer[Title] OR deers[Title] OR bear[Title] OR bears[Title] OR panda[Title] OR
pandas[Title] OR "wild hog"[Title] OR "wild boar"[Title] OR fitchew[Title] OR fitch[Title] OR beaver[Title] OR beavers[Title] OR jerboa[Title] OR jerboas[Title] OR capybara[Title] OR capybaras[Title] OR ovine[Title] OR
#13 ewe*[Title] OR chick*[Title] OR slug[Title] OR slugs[Title] OR pigeon[Title] OR pigeons[Title] OR dalmation*[Title] OR feline*[Title] OR bovine[Title]) 2442140
AC

#12 #5 OR #11 1746

#11 #1 AND #10 821
#10 #6 OR #9 9639
#9 #7 AND #8 6419
(((congenital[Title/Abstract]) OR aganglion*[Title/Abstract]) OR congenital[Other Term]) OR
#8 aganglion*[Other Term] 203043
(((((((megacolon[Title/Abstract]) OR colon*[Title/Abstract]) OR rectosigmoid[Title/Abstract]) OR
intestin*[Title/Abstract]) OR megacolon[Other Term]) OR colon*[Other Term]) OR
#7 rectosigmoid[Other Term]) OR intestin*[Other Term] 633561
#6 (hirschsprung*[Title/Abstract]) OR hirschsprung*[Other Term] 4441
#5 #1 AND #4 1091
#4 #2 AND #3 11390
(((((((((((((((((((((((artificial*[Title/Abstract]) OR malformation[Title/Abstract]) OR
anomal*[Title/Abstract]) OR abnormal*[Title/Abstract]) OR ectopic[Title/Abstract]) OR
stenosis[Title/Abstract]) OR atres*[Title/Abstract]) OR atroph*[Title/Abstract]) OR
imperforat*[Title/Abstract]) OR inperforat*[Title/Abstract]) OR praet*[Title/Abstract]) OR
pret*[Title/Abstract]) OR artificial*[Other Term]) OR malformation[Other Term]) OR anomal*[Other
#3 Term]) OR abnormal*[Other Term]) OR ectopic[Other Term]) OR stenosis[Other Term]) OR 1183079
 ACCEPTED MANUSCRIPT

atres*[Other Term]) OR atroph*[Other Term]) OR imperforat*[Other Term]) OR inperforat*[Other

Term]) OR praet*[Other Term]) OR pret*[Other Term]
(((((((anal[Title/Abstract]) OR anus[Title/Abstract]) OR anorect*[Title/Abstract]) OR
rectal[Title/Abstract]) OR anal[Other Term]) OR anus[Other Term]) OR anorect*[Other Term]) OR
#2 rectal[Other Term] 106232
(((((((((((((((ostom*[Title/Abstract]) OR stoma*[Title/Abstract]) OR stomy*[Title/Abstract]) OR

T
colostom*[Title/Abstract]) OR ileostom*[Title/Abstract]) OR cecostom*[Title]) OR
coecostom*[Title/Abstract]) OR caecostom*[Title/Abstract]) OR ostom*[Title/Abstract]) OR

P
stoma*[Other Term]) OR stomy*[Other Term]) OR colostom*[Other Term]) OR ileostom*[Other
#1 Term]) OR cecostom*[Other Term]) OR coecostom*[Other Term]) OR caecostom*[Other Term] 156284

RI
SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
SC
NU
MA
ED
PT
CE
AC

Figure 1
 ACCEPTED MANUSCRIPT

PT
RI
SC
Figure 2a

NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
SC
NU
MA
ED

Figure 2b
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
Figure 3

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
Figure 4

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
Figure 5

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
Figure 6

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
Figure 7

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
Figure 8

SC
NU
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

PT
RI
SC
NU
MA
ED
PT
CE
AC