Background: There is a paucity of current multicenter data assessing the outcomes of neonates with esophageal atresia and tracheoesophageal fistula (EA/TEF). This study aims to quantify mortality rates and hospital lengths of stay (LOS) in neonates with EA/TEF as well as to characterize the effects of birthweight (BW) and associated congenital anomalies upon these. Methods: Data regarding patients with EA/TEF were prospectively collected over a six-year period (2013-2019) at 298 North American centers. Survival was defined as discharge home or alive in hospital at 1 year of age. Infants who died at delivery were included. The primary outcome was mortality and secondary outcome was LOS. Risk factors affecting both mortality and LOS were assessed via multivariate regression using a generalized estimating equation to control for clustering of infants within hospitals. Poisson distribution was used in the analysis for mortality and a negative binomial distribution was used for LOS. Results: EA/TEF was diagnosed in 3290 infants with a median BW of 2476g (IQR 1897, 2970) and 12.8% had a BW <1500g. Over one-third of the cohort had a significant congenital anomaly other than EA/TEF (36.6%), of which severe cardiac (19%), gastrointestinal (GI) (18%), and chromosomal anomalies (10.2%) were the most common. In-hospital mortality was 12.7%. Mortality for patients with a BW >2500g was 3.8% compared to 43.5% in those patients with a BW <1500g, and was overall inversely correlated with BW (Figure 1). Based on multivariate regression, with every 100g increase in BW, the risk of mortality decreased by approximately 10% [RR 0.90 (95%CI 0.89, 0.92)]. Risk of mortality increased most notably for infants with EA/TEF who also had a severe cardiac anomaly [RR 7.81 (95%CI 6.09, 10.02)], GI anomaly [7.47 (5.39, 10.36)], or chromosomal anomaly [5.42 (4.12, 7.12)] when compared to infants without an anomaly in addition to EA/TEF (Table 1). Lower BW was associated with an increased mean LOS among survivors. Adjusted LOS for BW >2500g was 44 days (95%CI 42, 47), while adjusted LOS for infants with BW <1500g was 132 days (120, 146). Patients with a severe cardiac anomaly had an adjusted LOS of 106 days (91, 123) compared to 75 days (56, 100) for GI, and 79 days (69, 90) for chromosomal anomalies. Conclusion: This study, including more than 3,000 North American infants with EA/TEF, demonstrates a striking in-hospital mortality of over 10%. Both increased mortality and prolonged LOS are highly associated with lower birthweight and the presence of concomitant congenital anomalies. Awareness of these risk factors can assist surgeons in accurately determining prognosis and anticipating healthcare resource utilization for neonates with EA/TEF.
Predictors of mortality in patients with EA/TEF: Generalized estimating equation with Poisson distribution using log function to estimate risk ratios for mortality are reported. The reference cohort for birthweight (BW) consisted of infants with BW >2500g. The reference cohort for the congenital anomaly groups consisted of infants without additional severe congenital anomalies.