Anastomotic Leak and its implications: a multicenter analysis of Type C esophageal atresia TEF

Ankur Datta, Karna Murthy, Isabella Zaniletti, Yigit Guner , Michael A Padula, Theresa R Grover, Benjamin Zendejas, Shawn D St Peter,Jose Diaz-Miron, Mark Speziale, Jacquelyn R Evans, Loren Berman

Purpose

Repair of type C esophageal atresia with tracheo-esophageal fistula (EA/TEF) may be complicated by esophageal anastomotic leak. Risk factors associated with leak and the impact of leak on inpatient outcomes remains uncertain. Our objectives are to estimate the associations between clinical factors and esophageal anastomotic leak and quantify the association of leak with length of stay (LOS) in infants who underwent repair of type C EA/TEF.

Methods

Using the Children’s Hospitals Neonatal Database (CHND), we identified infants with type C EA/TEF from 2021 to 2023. The main outcomes were anastomotic leak and LOS. Multivariable associations between patient and clinical factors and these outcomes were quantified using logistic regression (leak) and Cox proportional hazards modelling (LOS).

Results

Among 365 infants at 36 centers, anastomotic leak occurred in 55 (15.1 %) infants, and thoracoscopic approach, lower birthweight, small for gestational age less than 10th percentile, male sex, staged repair, ventricular septal defect, and center were independently associated with leak (area under receiver operating curve = 0.853). Also, LOS was increased in infants with leak compared to those without [hazard ratio (HR): 0.655, 95 % CI = 0.431–0.996, p = 0.044], independent of birth weight, surgical approach, male sex, or VSD. The adjusted LOS demonstrated a 11-fold inter-center variation (p = 0.034).

Conclusions

Several clinical and operative factors are associated with esophageal anastomotic leak in infants after type C EA/TEF repair. Leak significantly prolongs LOS. The magnitude of inter-center variability in LOS also suggests that identifying best practices could aid in improving patient care in this patient population.