Journal of Pediatric Surgery 50 (2015) 107110

Contents lists available at ScienceDirect

Journal of Pediatric Surgery

journal homepage: www.elsevier.com/locate/jpedsurg

The Esophageal Anastomotic Stricture Index (EASI) for the management

of esophageal atresia
Linda Yi-Chan Sun a, Jean-Martin Laberge b, Yasmine Yousef a, Robert Baird b,
a
b

McGill University, Montreal, QC, Canada

Division of Pediatric Surgery, Montreal Children's Hospital; Faculty of Medicine, McGill University, Montreal, QC, Canada

a r t i c l e

i n f o

Article history:
Received 28 September 2014
Accepted 6 October 2014
Key words:
Esophageal atresia
Anastomotic stricture
Index
Prognostic tool

a b s t r a c t
Background: Anastomotic stricture is the most common complication following repair of esophageal atresia. An
Esophageal Anastomotic Stricture Index (EASI) based on the postoperative esophagram may identify patients
at high risk of stricture formation.
Methods: Digital images of early postoperative esophagrams of patients undergoing EA repair from 2005 to 2013
were assessed. Demographics and outcomes including dilations were prospectively collected. Upper (U-EASI)
and lower (L-EASI) pouch ratios were generated using stricture diameter divided by maximal respective
pouch diameter. Score performances were evaluated with area under the receiver operator curves (AUC) and
the Fisher's exact test for single and multiple (N3) dilatations. Interrater agreement was evaluated using the
intraclass correlation coefcient (ICC).
Results: Forty-ve patients had esophagrams analyzed; 28 (62%) required dilatation and 19 received N 3
(42%). U-EASI and L-EASI ratios ranged from 0.17 to 0.70, with L-EASI outperforming the U-EASI as follows:
L-EASI AUC: 0.66 for a single dilatation, 0.65 for N 3 dilatations; U-EASI AUC: 0.56 for a single dilatation, 0.67
for N3 dilatations. All patients with an L-EASI ratio of 0.30 (n = 8) required multiple esophageal dilatations,
p = 0.0006. The interrater ICC was 0.87.
Conclusion: The EASI is a simple, reproducible tool to predict the development and severity of anastomotic stricture after esophageal atresia repair and can direct postoperative surveillance.
2015 Elsevier Inc. All rights reserved.

Although the survival rate of patients with esophageal atresia (EA)

has improved drastically since the original description by Haight nearly
three quarters of a century ago, postoperative complications continue to
plague the early recovery of these neonates [1,2]. Anastomotic stricture
(AS) remains the most common complication following operative repair, and occurs in 20%50% of cases [3]. While some early predictors
of AS formation have been identied (long gap EA, esophageal leak),
the majority of patients continue to manifest this complication without
discernable risk factors. In the absence of these factors, clinicians should
aggressively treat gastroesophageal reux and investigate any
concerning symptomatology with vigilance.
Currently, a reliable prognostic tool to risk-stratify patients after
esophageal atresia repair has not been validated. The aim of this study
is to characterize such a prognostic tool based on the early postoperative esophagram. Such a tool would ideally form part of the standard
postoperative assessment after EA repair and demonstrate favorable
positive and negative predictive values for the development of a clinically signicant stricture. It should also prove easily applicable and reproducible. An objective Esophageal Anastomotic Stricture Index
(EASI) would allow meaningful comparisons of patient treatments
Corresponding author at: The Montreal Children's Hospital, 2300 Tupper Street, Room
C812, Montreal H3H 1P3, Quebec, Canada. Tel.: +1 514 412 4438; fax: +1 514 412 4289.
E-mail address: robert.baird@mail.mcgill.ca (R. Baird).
http://dx.doi.org/10.1016/j.jpedsurg.2014.10.008
0022-3468/ 2015 Elsevier Inc. All rights reserved.

and outcomes within and between institutions, and may aid in identifying patients at high risk of stricture formation with the potential to improve patient surveillance and inform parental counseling.
1. Methods
With IRB approval (11-267-PED), a single institution retrospective
review was conducted on all patients treated within our esophageal
atresia multidisciplinary clinic between 2005 and 2013. Patients were
included for analysis if they had an early postoperative esophagram
available on the Picture Archives and Communication System (PACS)
through Intelerad Distributed Radiology Solutions (Montreal, Canada).
Patients were excluded from analysis if they failed to undergo early
postoperative imaging, did not have an esophageal anastomosis
(H-type stula), or had inadequate imaging to create stricture index ratios. As part of routine follow-up, demographic information and characteristics of these patients were collected, including, gender, birth
weight, EA subtype and associated anomalies. Patient outcomes were
tracked, including survivorship, number of endoscopic dilatations performed, need for medical adjuncts during dilatations and the need for
reoperative surgery.
The EASI was generated after uoroscopic evaluation of the upper
gastrointestinal (GI) tract in the early postoperative period (postoperative days 510) using digital images that could be enlarged and

108

L.Y.-C. Sun et al. / Journal of Pediatric Surgery 50 (2015) 107110

manipulated as needed to obtain ratios. Two pouch ratios were generated by using the narrowest stricture diameter well lled with contrast,
divided by the maximal upper and lower respective pouch diameter
for the upper (U-EASI) and lower (L-EASI). Anteroposterior and lateral
projections were averaged to obtain ratios with equal weight on each
view for the U-EASI and L-EASI as demonstrated in the following
equations, (Fig. 1):

UEASI :

Lateral d=D Anteroposterior d=D

2
where d stricture diameter
D upper pouch diameter

LEASI : Lateral d=D Anteroposterior d=D

2
where d stricture diameter
D lower pouch diameter

A ratio of .25 means that the diameter at the anastomosis is 25% of

the diameter of the patient's normal esophagus. The smaller the ratio,
the tighter the stricture.
The relationship between the stricture ratio of the upper and lower
pouches and the number of dilatations needed post-EA surgery were
evaluated. We sought to determine whether cut-off values based on
these ratios could be correlated with need for dilatation. Statistical analyses of proposed cutoffs were carried out to conrm EASI ranges associated with dilatations beyond chance effects. Standard contingency table
association tests were created and analyzed using the Fisher's exact test
for single and multiple (N 3) dilatations, with a two-tailed level of significance of p b 0.05. Score performances were evaluated with area under
the receiver operator characteristic curves (AUC). The reproducibility of
the proposed tool was evaluated via intrarater and interrater agreement. Separate independent calculations of all EASI scores were performed by the same observer at two different times (intrarater),
followed by complete EASI determination by a distinct independent observer (interrater). Both measurements were characterized via
intraclass correlation coefcients (ICC).

Table 1
Demographic and outcome data for esophageal atresia patients included in the analysis.
Demographic data (n = 45)
Female gender, n (%)
Gestational age (median, range)
Birth weight (mean SD)
Atresia subtype
Type C, n (%)
Type D, n (%)
Type A, n (%)
Type B, n (%)
Outcome (n = 45)
Survival, n (%)
Length of stay (median, range)
Days of parenteral nutrition (median, range)

23 (51.1%)
38 weeks (2941)
3.12 kg 744.3 g
39 (87%)
3 (6.7%)
2 (4.4%)
1 (2.2%)
45 (100%)
14 days (8167)
7 days (3133)

and outcome data of the 45 included patients are presented in Table 1.

Twenty-eight (62%) patients had EA repair surgery on the second day
of life, while ve patients had EA repair on the day of birth (range
052 days). The median follow-up period for EA patients was
3.7 years (45 months11.2 years).
In total, there were 148 dilatations performed on 28 patients (62%),
ranging from 1 to 15 procedures per patient. Nineteen patients (42%)
received N3 dilatations. Stricture resolution occurred after a median of
5 dilatations (mean 5.29 3.65). Adjunct procedures (steroids and
mitomycin-C) were used in 18 patients, representing 64% of patients
that required dilatations.
The range of the stricture ratio using the rst postoperative
esophagram for the U-EASI is 0.120.52, and for L-EASI, it is 0.170.7.
The L-EASI, based on the lower pouch outperforms the U-EASI for
predicting the need for at least one dilatation as demonstrated by the
AUC as follows (Fig. 2): L-EASI AUC: 0.66; U-EASI AUC: 0.56. Areas
under the receiver operating characteristic curves were comparable
when evaluating patients requiring multiple (N 3 dilatations) L-EASI:
0.65, U-EASI: 0.67, (receiver operating characteristic curves not
shown). All patients with an L-EASI ratio of 0.30 (n = 8) required multiple esophageal dilatations (p = 0.0006; see Tables 2 and 3) while patients with an L-EASI ratio of 0.7 did not require any dilations (n = 2).
Intrarater and interrater ICCs were 0.91 and 0.89, respectively.
3. Discussion

2. Results
Of 72 patients EA patients evaluated over the study period, forty-ve
were ultimately included in the analysis. Four patients were excluded
because they had an H-type stula, 9 failed to undergo an early postoperative esophagram and 11 patients had early esophagrams that
predated their acquisition on digital media. Three patients had inadequate digital imagery for interpretation of EASI ratios. Demographic

More than half of EA patients who undergo surgical repair will develop signicant postoperative complications, with anastomotic stricture being the most common [3]. No reliable and widely-used tool has
been described to predict the development of AS after EA repair. The
predisposing factors for the formation of anastomotic stricture have
been identied as anastomotic leakage, anastomotic tension, increased
length of esophageal gap between the upper and lower pouch, suture

Fig. 1. Esophageal Atresia Stricture index (EASI) measurement. Note that the ratio is unitless and calculated automatically when measuring d/D, where d = stricture diameter
and D = diameter of esophageal pouch.

L.Y.-C. Sun et al. / Journal of Pediatric Surgery 50 (2015) 107110

109

AUC= area under the curve

Fig. 2. Receiver operator characteristic curves of the Esophageal Atresia Stricture Index (EASI) based on the upper (U-EASI) and lower pouch (L-EASI).

material and gastroesophageal reux disease (GERD) [47]. Routine

postoperative usage of H2 blockers and/or proton pump inhibitors is
strongly recommended, even though there is no proof of their efcacy
in preventing strictures [3]. The treatment of choice for patients with
AS after esophageal atresia repair is dilation, typically performed in the
symptomatic patient with radiologic evidence of stricture. Prophylactic
dilatations of the anastomosis have not shown any benet in the prevention of symptomatic strictures [8]. Physicians are therefore left with only
clinical symptoms to detect stricture formation, usually in a patient who
is already discharged from the hospital. Difculty bottling, coughing and
choking during feeding, regurgitation and apneic spells may be caused by
anastomotic stricture but may also be caused by a recurrent stula, severe tracheomalacia, GERD, swallowing incoordination or laryngeal
cleft. Such symptoms warrant initial investigation with a contrast
esophagram, followed as needed by exible upper esophagoscopy, rigid
or exible tracheoscopy, or a combination of these. At the time of symptomatic presentation, a high-grade stricture is often discovered. Currently
there are no recommendations to perform a routine delayed esophagram
(i.e. 46 weeks postoperatively), in part because of the low yield and the
fear of unnecessary radiation exposure. The ability to predict which patient will develop an anastomotic stricture (AS) would allow a more
timely management and may improve the efcacy of treatment.
Previous publications have attempted to investigate anastomotic
strictures after EA repair. Said et al [9] proposed a stricture index,
SI = A a/A, where A is the esophageal diameter of lower pouch and
a is the stricture diameter. This publication evaluated the utility of balloon dilatation under uoroscopy, but did not attempt to validate their
stricture index. Its publication has not proven impactful, having been
accurately cited only once [10]. Other investigators have quoted Said,
but incorrectly transformed the index to the endoscopic assessment of
the stricture and the upper pouch diameter [11]. Parolini et al [11] did
not give details on how the diameter was measured and did not comment on reproducibility. They also noted that a varying degree of physiological stenosis can be found on the rst esophagram as a normal
healing process from the surgical procedure, preferring to calculate
stricture index after 1 month of surgery. However, most surgeons perform a baseline esophagram 510 days postoperatively to eliminate

an anastomotic leak. Nambirajan et al [12] also used the upper pouch diameter, only as measured during contrast esophagram.
The EASI is a simpler formulation (d/D), and we considered both the
lower and upper pouches independently in our validation. Our data
demonstrates that the lower pouch stricture ratio, L-EASI, is superior
in determining the prognosis for an EA patient. An L-EASI of 0.30 is
highly correlated with requiring a course of dilatation. This dichotomization allows for identifying patients at increased risk of AS; surveillance and counseling can be tailored accordingly. The range for L-EASI
(0.170.7) was larger than that of the U-EASI (0.120.52). It was observed on the esophagrams that the upper pouch is consistently dilated,
likely caused by chronic obstruction in utero. The lower pouch therefore
provides a more accurate reection of the true esophageal diameter,
and should form the basis of the EASI. Should the EASI be 0.3, the family should be counseled about the high risk of AS and its signs, and a repeat esophagram could be done at the rst appearance of symptoms.
Early identication of an AS increases the likelihood of treating the stricture prior to the development of dense brosis [9]. Some may wish to
forfeit any repeat esophagram to reduce exposure to ionizing radiation
and proceed with esophagoscopy and dilatation under general anesthesia GA, but the risks of affecting neurocognitive developmental from repeat GA in infants have to be taken into consideration as well [13].
Anastomotic stricture after EA repair is common. The intent of EASI
is to create an easy, reproducible tool to predict the development and
severity of AS that pediatric surgeons and gastroenterologists can use
to identify patients at risk. Incorporation into current practice patterns
appears straightforward. Baird et al [3] proposed an algorithm for the
management of AS (Fig. 3), which has been modied to incorporate
the EASI evaluation during the postoperative esophagram. This tool
can guide the frequency of follow-up visits as well as the scheduling
of contrast studies or upper endoscopy to investigate symptoms. The
EASI can also be used in the future to correlate the severity of strictures
with the efcacy of various treatment methods (dilatation interval,
intralesional steroids, topical Mitomycin C, stents, etc.). The EASI may
also be used in the future to compare anastomotic techniques in patient
registries such as EUROCAT [14]. Ideally, it should be further evaluated

Table 2
Contingency table of Lower Esophageal Atresia Stricture Index (L-EASI) on the need for dilatation after esophageal atresia repair using 0.3 as a cut-point.

Table 3
Contingency table of Lower Esophageal Atresia Stricture (L-EAS) on the need for multiple
dilatations after esophageal atresia repair using 0.3 as a cut-point.

Stricture ratio

Needed dilatation

Did not need dilatation

0.30
N0.30
Total

8
20
28

0
17
17

p = 0.02

Stricture ratio

N3 dilatations

3 dilatations

0.30
N0.30
Total

8
12
19

0
25
26

p = 0.0006

110

L.Y.-C. Sun et al. / Journal of Pediatric Surgery 50 (2015) 107110

Fig. 3. An algorithm for management of esophageal anastomotic stricture adapted from by Baird et al. incorporating the Esophageal Atresia Stricture Index (EASI).

prospectively in a large cohort of patients from different centers to

validate its usefulness and reproducibility.
References
[1] Alshehri A, Lo A, Baird R. An analysis of early nonmortality outcome prediction in
esophageal atresia. J Pediatr Surg 2012;47:8814.
[2] Konkin DE, O'hali WA, Webber EM, et al. Outcomes in esophageal atresia and
tracheoesophageal stula. J Pediatr Surg 2003;38:17269.
[3] Baird R, Laberge J-M, Lvesque D. Anastomotic stricture after esophageal atresia repair: a critical review of recent literature. J Pediatr Surg 2013;23:20413.
[4] Antoniou D, Soutis M, Christopoulos-Geroulanos G. Anastomotic strictures following
esophageal atresia repair: a 20-year experience with endoscopic balloon dilatation.
J Pediatr Gastroenterol Nutr 2010;51:46474.
[5] Serhal L, Gottrand F, Sfeir R. Anastomotic stricture after surgical repair of esophageal
atresia: frequency, risk factors, and efcacy of esophageal bougie dilatations.
J Pediatr Surg 2010;45:145962.
[6] Chittmittrapap S, Spitz L, Kiely EM, et al. Anastomotic leakage following surgery for
esophageal atresia. J Pediatr Surg 1992;27:2932.

[7] Michaud L, Gottrand F. Anastomotic strictures: conservative treatment. J Pediatr

Gastroenterol Nutr 2011;52:189.
[8] Koivusalo A, Pakarinen MP, Rintala RJ. Anastomotic dilatation after repair of esophageal atresia with distal stula. Comparison of results after routine versus selective
dilatation. Dis Esophagus 2009;22:90194.
[9] Said M, Mekki M, Golli M, et al. Balloon dilatation of anastomotic strictures secondary to surgical repair of oesophageal atresia. Br J Radiol 2003;76:263115.
[10] Zhao R, Li K, Shen C, et al. The outcome of conservative treatment for anastomotic
leakage after surgical repair of esophageal atresia. J Pediatr Surg 2011;46(12):
22748 [18].
[11] Parolini F, Leva E, Morandi A, et al. Anastomotic strictures and endoscopic dilatations
following esophageal atresia repair. Pediatr Surg Int 2013;29:6015.
[12] Nambirajan L, Rintala RJ, Losty PD, et al. The value of early postoperative
oesophagography following repair of oesophageal atresia. Pediatr Surg Int 1998;
13:768.
[13] Sun L. Early childhood general anaesthesia exposure and neurocognitive development. Br J Anaesth 2010;105(Suppl 1):i618.
[14] Pedersen RN, Calzolari E, Husby S, et al. Oesophageal atresia: prevalence, prenatal diagnosis and associated anomalies in 23 European regions. Arch Dis Child 2012;97
(3):22732.