pennington 2014

8/15/2019 Pennington 2014

1/6

Use of mechanical bowel preparation and oral antibiotics for elective

colorectal procedures in children: Is current practice evidence-based?

Elliot C. Pennington a, Christina Feng a, Shawn D. St. Peter b, Saleem Islam c, Adam B. Goldin d,Fizan Abdullah e, Shawn J. Rangel a,⁎a Boston Children’ s Hospital and Harvard Medical School, Boston, Massachusettsb Children's Mercy Hospitals & Clinics, Kansas City, Missouric University of Florida College of Medicine, Gainesville, Floridad Seattle Children's Hospital, Seattle, Washingtone Johns Hopkins University, Baltimore, Maryland

a b s t r a c ta r t i c l e i n f o

Article history:

Received 26 January 2014

Accepted 27 January 2014

Key words:

Mechanical bowel preparation

Non-absorbable oral antibioticsl

colorectal surgery

Antibiotic prophylaxis

Purpose: It is well established through randomized trials that oral antibiotics given with or without a

mechanical bowel preparation (MBP) prior to colorectal procedures reduce complications, while MBP given

alone provides no benet. We aimed to characterize trends surrounding bowel preparation in children and

determine whether contemporary practice is evidence-based.

Methods: Retrospective analysis of patients undergoing colorectal procedures at 42 children ’s hospitals (1/2/

2007-12/31/2011) was performed. Patients were analyzed for diagnosis, pre-admission status, and inpatient

bowel preparation. Bowel preparation was considered evidence-based if oral antibiotics were utilized with or

without a MBP.

Results: 49% of all patients were pre-admitted (n = 5,473), and the most common diagnoses were anorectal

malformations (55%), inammatory bowel disease (26%), and Hirschsprung’s Disease (19%). The most

common preparation approaches were MBP alone (54.3%), MBP + oral antibiotics (18.8%), and oral

antibiotics alone (4.2%), although signicant variation was found in hospital-specic rates for each approach

(MBP alone: 0-96.1%, MBP + oral antibiotics: 0-83.6%, orals alone: 0-91.6%, p b 0.0001). Only 22.9% of all

patients received an evidence-based preparation (range by hospital: 0-92.3%, p b 0.0001), and this ratedecreased signicantly during the ve-year study period (27.6% in 2007 vs. 17.3% in 2011, p b 0.0001).

Conclusion: According to the best available clinical evidence, less than a quarter of all children pre-admitted

for elective colorectal procedures receive a bowel preparation proven to reduce infectious complications.

© 2014 Elsevier Inc. All rights reserved.

The rates of infectious complications associated with colorectal

surgery are among the highest with respect to elective procedures

performed in adults and children alike. Postoperative complications

for patients with inammatory bowel disease occur in up to 55% of

patients, and patients undergoing colorectal operations for Hirsch-

sprung’s disease may experience a complication in close to 40% of

cases [1,2]. Given the relatively high burden of infectious complica-

tions attributable to colorectal procedures, many prophylactic

strategies have been utilized to mitigate these events in the

perioperative setting. Intravenous antibiotics at the time of surgery

are widely considered the standard of care in this regard, however

opinions remain mixed about the role of mechanical bowel prepara-

tion (MBP) and oral antibiotics as prophylactic adjuncts to further

decrease the risk of complications.

The published evidence examining the prophylactic effectiveness

of MBP and non-absorbable oral antibiotics in adult colorectal surgery

is fairly extensive. Several meta-analyses incorporating data from

high-quality randomized controlled trials havermly established that

non-absorbable oral antibiotics used with or without a MBP

signicantly reduce complications, while the administration of a

MBP alone (without oral antibiotics) provides no benet. Although no

published trial has compared oral antibiotics alone with oral

antibiotics combined with a MBP, data from two large, prospective,

multi-center colorectal outcomes databases have suggested that oral

antibiotics combined with a MBP signicantly reduce infectious

complications compared with MBP alone [3,4]. From this pool of

relatively high-quality clinical evidence, two conclusions surrounding

colorectal prophylaxis can be reached: 1) Oral non-absorbable

antibiotics should be administered (with or without a MBP) as a

prophylactic adjunct to intravenous antibiotics for elective colorectal

procedures, and 2) MBP should never be given alone (without oral

antibiotics) for this purpose.

Journal of Pediatric Surgery 49 (2014) 1030–1035

⁎ Corresponding author at: Boston Children’s Hospital 300 Longwood Avenue, Fegan

3 Boston, MA 02115. Tel.: + 1 617 355 3040; fax: +1 617 730 0752.

E-mail address: [email protected] (S.J. Rangel).

http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048

0022-3468/© 2014 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Journal of Pediatric Surgery

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j p e d s u r g

http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048mailto:[email protected]://dx.doi.org/10.1016/j.jpedsurg.2014.01.048http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048http://www.sciencedirect.com/science/journal/http://www.sciencedirect.com/science/journal/http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048http://dx.doi.org/10.1016/j.jpedsurg.2014.01.048mailto:[email protected]://dx.doi.org/10.1016/j.jpedsurg.2014.01.048http://crossmark.crossref.org/dialog/?doi=10.1016/j.jpedsurg.2014.01.048&domain=pdf


2/6

With these considerations, the aims of this study were twofold: 1)

To characterize overall and hospital-specic variation in pre-admis-

sion and the use of different bowel preparation strategies for

commonly performed colorectal procedures in children, and 2) To

assess overall and hospital-specic variation in the utilization of

bowel preparation strategies that arecurrently supportedby thehigh-

quality evidence cited above.

1. Methods

After obtaining IRB approval, we performed a 5-year (1/2/2007-

12/31/2011) retrospective cohort study of all patients aged 3 months

to 18 years, admitted from the outpatient setting who underwent an

elective colorectal procedure at 42 children’s hospitals using the

Pediatric Health Information System (PHIS) database. Hospitals that

provide PHIS data are members of the Children’s Hospital Association

(Shawnee Mission, KS), and include freestanding, non-competing

children’s hospitals across the United States. Patient-level data

available through the PHIS database includes the following: demo-

graphic and payer information; primary and secondary International

Classi cation of Diseases, Ninth Revision, Clinical Modi cation (ICD-9-

CM) diagnostic and procedural codes; and date-stamped billing data

for a range of diagnostic tests, therapeutic procedures, and adminis-

tration of pharmacologic agents.

1.1. Cohort de nitions

Patients were identied using a search algorithm of 40 ICD-9-CM

colorectal procedural codes used to identify colorectal procedures.

These included 20 pullthrough or anorectoplasty procedures involving

the colon, rectum or anus, 17 procedures involving colon resections

with or without a colo-enteric anastomosis, and 3 procedures for

colostomy creation, revision or closure. Cases where other procedures

were performed concomitantly on the esophagus, stomach, or

urogenital system were excluded as this may have impacted the

decision to utilize a bowel preparation on the basis of enteral tolerance

or for an indication other than colorectal prophylaxis.

Pre-admission was dened as a pre-operative hospital stay of 1-2calendar days. Patients who had an emergency department (ED) charge

during the same admissionwere excluded in order to limit the cohort to

only those electively admitted from the ambulatory setting. The

medicationprolesof allpatientswho were pre-admittedwere assessed

for Poly-ethylene Glycol (PEG) and Magnesium Citrate-based enteral

solutions, Eryrthomycin-base, and Neomycin Sulfate. Patients receiving

eitherof the enteral solutions were considered as havingreceiveda MBP

and patients receiving either (or both) Erythromycin base or Neomycin

Sulfate were considered having received oral antibiotics.

1.2. Analysis

The cohort was analyzed for demographic information, pre-

operative diagnosis, pre-admission status, and the type of bowelpreparation used. Rates of pre-admission, use of different bowel

preparation strategies, and utilization of evidence-based bowel

preparations were calculated for the entire cohort and for each

hospital. Given the strength and homogeneity of available evidence

from adult patients, we considered a patient to have received an

evidence-based bowel preparation if they received any oral antibiotics

(Erythromicin base or Neomycin Sulfate) with or without a

mechanical bowel preparation. Chi-square analysis was used to

compare proportions between groups and the Cochran-Armitage

test for trend was used to compare rates over time. All statistical

analysis was performed using SAS v9.3 (SAS Institute Inc., Cary, NC).

2. Results

A total of 7,056 patients underwent an elective colorectal

procedure during the study period. The most common diagnoses

included anorectal malformations (ARM) in 42.6% (3005/7056),

inammatory bowel disease (IBD) in 20.1% (1417/7056), and

Hirschsprung’s Disease (HD) in 14.9% (1051/7056). Within the cohort

of patients with IBD, 42.3% had ulcerative colitis, 39.3% had Crohn ’s

Disease, and 18.4% had mixed colitis. Of note, 21.4% of the cohort

(1,513/7056) did not have a preoperative diagnosis identifying a

specic colorectal condition. As it was our intent to examine practice

variationin the context of specic pediatricdiseases, we chose to limit

our analysis on the three most common conditions associated with

elective colorectal procedures in children (ARM’s, HD and IBD).

Demographic data for this sub-cohort of children (n = 5,473) is

presented in Table 1. The median age was 17 months (IQR: 6-

137 months), and 53.8% were male. The median number of patients

per hospital was 116.5 (range: 21-774).

2.1. Pre-admission rates

The overall rate of pre-admission for the entire cohort was 47.5%

(2,599/5,473) and this varied signicantly by pre-operative diagnosis

(HD: 63.8%; ARM: 46.2%; IBD: 38.2%; p b 0.0001). Of the pre-admitted

cohort, 43% had a preoperative length of stay of one day and 4.5% two

days. A 12-fold variation in pre-admission rates was observed across

hospitals (range: 7-86%, p b 0.0001), and the magnitude of thisvariation was even greater when analyzed by individual diagnosis

(ARM range: 0-91.2; IBD range: 0-90.0%; HD range: 0-100%,

p b 0.0001 for each diagnosis, Fig. 1-C).

2.2. Bowel preparation strategies

In the pre-admitted cohort, 54.3% (n = 1,412)received PEG alone,

18.8% (n = 488) received PEG along with oral antibiotics, 4.2% (n =

108) received oral antibiotics alone, and 22.7% (n = 591) received

neither PEG nor oral antibiotics. Fifty-seven (2.2%) patients received

an electrolyte laxative solution (e.g. Magnesium Citrate). This relative

preference for bowel preparation approach was similar for each of the

three diagnoses. Of the children receiving any oral antibiotics with or

without a MBP (n = 596), 46.8% received Neomycin Sulfate as theonly oral antibiotic, 3.9% received Erythromycin base as the only oral

Table 1

Demographic information, pre-admission rates, and bowel preparation strategies for patients undergoing elective colorectal procedures at 42 children ’s hospitals 2007-2011.

Pre-operative diagnosis n (%) Median age in months (+/- IQR) Male (%) Preadmission rate (%) Preadmission bowel preparation strategy

MBP (%) Orals (%) MBP + Orals (%)

Hirschsprung’s Disease 1051 (19.2) 12 (6-48) 74.7 63.8 59.7 3.9 21.2

Ano-rectal Malformations 3005 (54.9) 9 (5-20) 48.0 46.2 54.2 4.3 14.6

Inammatory Bowel Disease 1417 (25.9) 183 (152-205) 50.8 38.2 48.0 4.2 26.4

Ulcerative Colitis 603 (11.0) 178 (138-201) 49.6 41.5 52.4 3.6 26.0

Crohn’s Disease 554 (10.1) 192 (169-213) 50.2 32.9 43.4 3.9 24.7

Mixed Colitis 260 (4.8) 175 (150-197) 54.6 42.3 45.5 6.4 30.0

All 5473 (100) 17 (6-137) 53.8 47.4 54.3 4.2 18.8

1031E.C. Pennington et al. / Journal of Pediatric Surgery 49 (2014) 1030–1035


3/6

Fig. 1. A Variation in pre-admission rates and associated bowel preparation strategy in patients undergoing elective colorectal procedures for ano-rectal malformations at 42

children's hospitals. B Variation in pre-admission rates and associated bowel preparation strategy in patients undergoing elective colorectal procedures for In ammatory Bowel

Disease at 42 children's hospitals. C Variation in pre-admission rates and associated bowel preparation strategy in patients undergoing elective colorectal procedures for

Hirschsprung’s Disease at 42 children's hospitals.

1032 E.C. Pennington et al. / Journal of Pediatric Surgery 49 (2014) 1030–1035


4/6

antibiotic, and 49.4% received both agents 90.1% (n = 2365) received

intravenous antibiotics at the time of surgery.

The rates of different bowel preparation approaches varied

signicantly across hospitals (PEG alone: 0-96.1%; PEG + oral

antibiotics: 0-83.6%; oral antibiotics alone: 0-91.6%, p b .0001 for

each approach, Fig. 1-C). When analyzed by individual diagnosis,

signicant variability between hospitals persisted although PEG alone

remained the most common preparation strategy used.

2.3. Evidence-based practice

The overall utilization rate of a bowel preparation approachsupported by the best available adult evidence shown to reduce

infectious complications was 22.9% (596/2599), and this rate varied

signicantly among hospitals (range: 0-92.3%, p b 0.0001, Fig. 2). No

patient received an evidence-based preparation in 13 (31%) of the 42

hospitals examined during the ve-year study period.

2.4. Trends over time

The rate of pre-admission decreased signicantly during the study

period, (51.6% in 2007 vs. 41.4% in 2011, p b 0.0001). The relative

proportion of patients receiving PEG alone increased by 42% (41.2% in

2007 vs. 59.0% in 2011, p b 0.0001), while the rates of all other

preparation strategies remained unchanged or decreased. The relative

proportion of patients receiving a bowel preparation approach shown

to reduce infectious complications in adult patients decreased by 37%

during the study period, from 27.6% in 2007 to 17.3% in 2011

(p b 0.0001, Fig. 3).

3. Discussion

In this study of 42 freestanding children’s hospitals, we found

signicant variation in the rates of pre-admission associated with

elective colorectal surgery and an equally striking variation in the

Fig. 2. Variation in the proportion of pre-admitted patients undergoing elective colorectal procedures who received an evidence-based bowel preparation treatment. (either oral

antibiotics only or oral antibiotics combined with a mechanical bowel preparation).

Fig. 3. Variation in pre-admission rates and associated bowel preparation strategy over time in patients undergoing elective colorectal procedures at 42 children's hospitals.


http://localhost/var/www/apps/conversion/tmp/scratch_3/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_3/image%20of%20Fig.%E0%B2%80


5/6

preferred approach to bowel preparation. Perhaps even more striking

was thending that the majority of children pre-admitted for elective

colorectal procedures received a bowel preparation shown to be

ineffective in high-quality clinical studies involving adult patients.

Furthermore, the relative proportion of children receiving a bowel

preparation that has been shown to reduce infectious complications

appears to be decreasing over time.

The reasons for the broad variation observed in our study (and

apparent poor compliance with evidence-based guidelines) are likely

complex and multi-factorial. One possibility is that pediatric surgeons

may not be familiar with the strength of evidence in support of oral

antibiotics, and conversely, against the routine use of MBP alone for

colorectal prophylaxis. In this regard, there exist many high-quality

randomized clinical trials in the adult literature comparing the

ef cacy of different bowel preparation strategies. The majority of

these have examined the use of mechanical bowel preparation (e.g.

PEG) compared to no preparation [5–8]. The most comprehensive of

these reviews, from the Cochrane Collaboration published in 2011,

included 18 randomized controlled trials with a total of 5,805

patients. Their analysis denitively showed that mechanical prepara-

tion as a sole adjunct to IV antibiotics at the time of surgery conferred

no benet compared to IV antibiotics alone [8]. Other meta-analyses

have come to the same conclusion [9,10].

The benet of oral non-absorbable antibiotics as an adjunct tointravenous antibiotics has also been extensively studied and

conrmed through a multitude of randomized trials. In a meta-

analysis of 16 such trials that included 2,669 patients, the addition

of oral non-absorbable antibiotics to standard parenteral antibiotics

at the time of surgery reduced the risk of surgical site infection by

43% compared with parenteral antibiotics alone [11]. There are

currently no published trials that have examined whether a MBP

should be combined with oral antibiotics to increase their ef cacy.

However, the results from two large, consortium-based multi-center

colorectal collaborative studies have suggested that the combination

of MBP and oral antibiotics reduces infectious complications by more

than half when compared with a mechanical preparation alone or no

preparation (neither a mechanical preparation or oral antibiotics)

[3,4]. Although neither of these studies was able to provide insightinto whether the ef cacy of oral antibiotics is affected with the

addition of a MBP, it would seem that both approaches are associated

with a signicant reduction in the risk infectious complications. Taken

collectively, the available evidence in the adult population would

strongly suggest that oral antibiotics should always be given as a

prophylactic adjunct (with or without a MBP), while mechanical

preparation alone should never be used for this purpose.

The marked variation in practice mayalso be explained in part by a

lack of equipoise among pediatric surgeons with the relevance of

existing data to the pediatric population. It is important to note that

the “high-quality” evidence cited above is entirely derived from the

adult colorectal experience. While the evidence is clear that mechan-

ical bowel preparation offers no benet in adults, the use of MBP and

non-absorbable oral antibiotics has not been as extensively studied inchildren. One recent multi-center retrospective review of children

undergoing colostomyclosureshowed that MBP actually increasedthe

risk of wound infection (14.4% vs. 5.8%) and length of stay (5.6 vs

4.4 days), while the addition of oral antibiotics had no effect [12].

Another study of a similar cohort of patients undergoing colostomy

closure forARM’s found no benet in the addition of oral antibiotics to

MBP in the rate of infectious complications (13% vs 17%) [13]. To date

there have been no prospective studies of any type comparing bowel

preparation strategies in the pediatric population. Given this limited

pediatric data, we are challenged with the question: Are the results

from high-quality, adult-focused studies generalizable to the pediatric

population?Althoughwe would alltend to agreethat “childrenare not

small adults” when it comes to colorectal diseases and their operative

management, the relative concentration and species of colonic

bacteria are similar between children and adults by the rst year of

life. This would suggest that interventions to reduce bacterial

concentration in the colon and rectum are likely relevant to all

patients undergoing colorectal procedures, young and old alike [14].

Thendings in this study must be carefully considered in the context

of its limitations. This study was based on the retrospective analysis of

administrative data and some degree of error (including miscoding,

misclassication, or omitted data) is inherent. In this regard, 22.7% of

patients admitted prior to surgery received neither a mechanical

preparation nor oral antibiotics. It was not possible to establish why

these patients were pre-admitted, although plausible reasons may

include social and travel considerations, preoperative diagnostic testing,

and anesthesia evaluation for patients with complex care conditions,

among others. Furthermore, some patients may have been admitted to

undergo a “modied” bowel preparation using a clear liquid diet. It is

also plausible that pediatric surgeons may utilize MBP for reasons other

than prophylaxis, such as decreasing the preoperative stool burden in

chronically constipated children (e.g. anorectal malformations). It could

be argued that MBP in these cases may make the operation technically

easier, reduce operative time and cumulative anesthetic exposure, and

even overall cost. While this has not been studied and would certainly

be a reasonable indication for using a MBP, there appears to be little

justication for omitting oral antibiotics as an additional adjunct given

their establishedprophylactic value. Finally,we werenot ableto examinebowel preparation practices in the cohort of patients that were not pre-

admitted. Many of these patients, especially those older patients with

IBD, may have received oral antibiotics at home resulting in an

underestimation of the overall compliance with the level-one evidence

derived from the adult literature.Wide variability was found among children’s hospitals in the

practice of pre-admission and bowel preparation strategies for

children undergoing elective colorectal procedures. The majority of

patients received bowel preparations that have not been shown to be

effective in reducing infectious complications, and this apparent gap

between the available evidence and clinical practice appears to be

growing over time. These ndings may have important implications

for pediatric surgical practice given the relatively high prevalence of

colorectal conditions in children, high cost associated with pre-admission, and quality of life considerations as nasogastric tubes are

frequently required for the administration of MBP’s. We owe it to our

patients to develop consensus guidelines around existing high-quality

evidence, and to develop the collaborative infrastructure to support

rigorous comparative effectiveness studies where equipoise remains

elusive and variation in care great.

References

[1] Jan S, Slap G, Dai D, et al. Variation in surgical outcomes for adolescents and youngadults with inammatory bowel disease. Pediatrics 2013;131(Suppl 1):S81–9.

[2] Yanchar NL, Soucy P. Long-term outcome after Hirschsprung's disease: patients'perspectives. J Pediatr Surg 1999;34:1152–60.

[3] Cannon JA, Altom LK, Deierhoi RJ, et al. Preoperative oral antibiotics reducesurgical site infection following elective colorectal resections. Dis Colon Rectum2012;55:1160–6.

[4] Englesbe MJ, Brooks L, Kubus J, et al. A statewide assessment of surgical siteinfection following colectomy: the role of oral antibiotics. Ann Surg2010;252:514–9 [discussion 519–520].

[5] Pineda CE, Shelton AA, Hernandez-BoussardT, et al. Mechanicalbowel preparationin intestinal surgery: a meta-analysis and review of the literature. J GastrointestSurg 2008;12:2037–44.

[6] Cao F, Li J, Li F. Mechanical bowel preparation for elective colorectal surgery:updated systematicreview andmeta-analysis.Int J ColorectalDis 2012;27:803–10.

[7] Slim K, Vicaut E, Launay-Savary MV, et al. Updated systematic review and meta-analysis of randomized clinical trials on the role of mechanical bowel preparationbefore colorectal surgery. Ann Surg 2009;249:203–9.

[8] Guenaga KF, Matos D, Wille-Jorgensen P. Mechanical bowel preparation forelective colorectal surgery. Cochrane Database Syst Rev 2011:CD001544.

[9] Guenaga KK, Matos D, Wille-Jorgensen P. Mechanical bowel preparation forelective colorectal surgery. Cochrane Database Syst Rev 2009:CD001544.

[10] Guenaga KF, Matos D, Castro AA, et al. Mechanical bowel preparation for elective

colorectal surgery. Cochrane Database Syst Rev 2005:CD001544.

1034 E.C. Pennington et al. / Journal of Pediatric Surgery 49 (2014) 1030–1035


6/6

[11] Bellows CF, Mills KT, Kelly TN, et al. Combination of oral non-absorbable andintravenous antibiotics versus intravenous antibiotics alone in the prevention of surgical site infections after colorectal surgery: a meta-analysis of randomizedcontrolled trials. Tech Coloproctol 2011;15:385–95.

[12] Serrurier K, Liu J, Breckler F, et al. A multicenter evaluation of the role of mechanical bowel preparation in pediatric colostomy takedown. J Pediatr Surg2012;47:190–3.

[13] Breckler FD, Rescorla FJ, Billmire DF. Wound infection after colostomy closure forimperforate anus in children: utility of preoperative oral antibiotics. J Pediatr Surg2010;45:1509–13.

[14] Palmer C, Bik EM, DiGiulio DB, et al. Development of the human infant intestinal

microbiota. PLoS Biol 2007;5:e177.

Discussion

MALE VOICE: - - from Houston. Thank you. That was a fantastic

summary, but I kind of was waiting for the punch line. Did you

make a difference?

DR. PENNINGTON: The study really isn't designed to say whether or

not it makes a difference, but based on the best available evidence,

which is primarily in adults, the use of oral antibiotics does make a

difference, and three-quarters of children in our series are not

getting that.

MALE VOICE: Again, back to your point, those studies were done in

adults. Children are not little adults. Yes, we probably should just

accept what adults do but in this case where you have this large

series that did not, it would have been interesting to see if you

made a difference or not.

DR. PENNINGTON: Sure. There is a small amount of data in children, all

retrospective reviews, mostly single study that sort of shows

conicting evidence about whether mechanical bowel prep and

oral antibiotics are as useful, but that remains to be studied and

we're involved in a large propensity-based analysis that will hope

to answer that question. So stay tuned.

DR. LOVVORN: Okay, thank you, Dr. Pennington.


pennington 2014

Documents