Facebook Twitter YouTube


Use of Preoperative Mechanical Ventilation, Birth Weight and Presence of Major Cardiac Anomalies to Improve Risk Stratification In Esophageal Atresia with or without Tracheo-Esophageal Fistula

Saturday, October 20, 2012: 1:52 PM
Napoleon Ballroom (Hilton Riverside)
Offir Ben-Ishay, Department of Surgery and the Advanced Fetal Care Center, Boston Children's Hospital, Boston, MA, Francesca R. Pluchinotta, MD, Department of Cardiology, Boston Children's Hospital, Boston, MA, Samuel C. Shecter, MD, Department of Surgery, University of California at San Francisco, San Francisco, CA, Ryan P. Cauley, MD, Surgery, Children's Hospital, Boston, Boston, MA, Wayne Tworetzky, MD, Pediatric Cardiology and The Advanced Fetal Care Center, Children's Hospital Boston, Boston, MA, Hanmin Lee, MD, FAAP, Surgery, Division of Pediatric surgery, UCSF Fetal Treeatment Center, University of California at San Francisco, San Francisco, CA, Anita J. Moon-Grady, MD, Pediatric Cardiology, University of California at San Francisco, San Francisco, CA and Terry L. Buchmiller, MD, FAAP, FACS, Department of Surgery, Advanced Fetal Care Center, Children's Hospital Boston, Boston, MA


Esophageal atresia (EA) with or without tracheo-esophageal fistula (TEF) continues to be associated with significant complications and long-term morbidity. Although low birth weight and major congenital heart disease (MCHD) have been associated with mortality (the Spitz classification), the discriminatory abilities of these two factors alone remain inadequate. We aimed to develop an enhanced classification system that is predictive of both mortality and secondary outcomes such as overall and ICU length of stay (LOS).


We performed a retrospective review of patients with EA first repaired at either of two participating academic children's hospitals from 1995-2011. 187 patients met inclusion criteria. The primary outcome measure was in-hospital mortality. Secondary outcomes included days on mechanical ventilation, ICU and total LOS, need for gastrostomy, and staged repair. Significant clinical predictors of outcomes were determined by univariate analysis. We developed a novel classification system incorporating significant crude predictors of mortality to maximize model discrimination of the studied outcomes (Figure 1).


There were 7 deaths (mortality 3.7%). Crude predictors of mortality included low birth weight (p=0.03), MCHD (p=0.005) and days on mechanical ventilation (p=0.02). Interestingly anastomotic leaks and stricture formation were not significantly associated with survival, low birth weight or MCHD. The Spitz classification system discriminated poorly between patients at moderate/high disease specific risk. Our new classification system creates 4 groups (A-D) according to weight (< or >1.5 kg), the presence or absence of MCHD, and POMV status (Figure 1). Univariate analysis comparing groups demonstrated a strong graded association with outcome (Table 1). Logistic and linear regression analysis controlling for surgical center, presence of long gap EA and APGAR score at 5 minutes, showed that our new classification predicts mortality (p=0.005), ICU (p=0.03) and hospital length of stay (p=0.001).


POMV, as a surrogate for the preoperative clinical condition of EA patients combined with birth weight and the presence or absence of MCHD, improves risk stratification and outcome prediction in contemporary EA patients.

Description: Description: Macintosh HD:Users:offirben:Desktop:Revised Classification.jpg

                             Revised Classification










p value

Mortality, n (%)

0 (0)

4 (10.3)

2 (11.1)

 1 (11.1)


Apgar (median)

     1 min (IQR)

 8 (6-8)

7 (5-8)

7 (7-8)

5 (3-6.5)


    5 min (IQR)

9 (8-9)

9 (7.5-9)

8 (8-9)

7 (6.5-7.5)


Staged EA repair (no large gap) (%)

2 (1.7)

4 (10.3)

2 (11.1)

5 (55.6)


Gastrostomy (no large gap) (%)

13 (10.7)

10 (25.6)

7 (38.9)

7 (77.8)


Mechanical ventilation, median (IQR) (d)

3 (1-6)

6.5 (3-25)

5 (3-21)

7 (5-33.5)


ICU stay, median (d)

13 (8-23)

25 (17.25-46.5)

35 (21-64)

44 (5-128)


Hospital Stay, median (IQR) (d)

19.5 (14-40)

39 (23-81)

53.5 (23.5-97.25)

102 (52.5-189.5)


EA – Esophageal Atresia, ICU – Intensive Care Unit