The effect of birth weight on survival
There were 81 cases of esophageal atresia and 24 cases of birth weight <2500g (30.9%), including 6 cases of very low birth weight <2000 g and 2 cases of extremely low birth weight <1500g (figure 1). In our study, 57 patients were born with normal birth weight. 5 died after surgery and the OS was 91.2%. OS was 75.9% in low-birth-weight infants, with statistical difference between the two groups (P=0.027). To analyze the data associated birth weight with survival rate, we found birth weight is positive correlation with survival rate (Pearson correlation coefficient of 0.936, P = 0.014, Fig 2). That is, the lower the birth weight, the lower the survival rate.
Among the 7 postoperative deaths of low-birth-weight infants, 1 case was long-segment esophageal atresia, which resulted in uncontrollable postoperative infection and pneumonia, and soon multi-system organ failure. The other 6 cases were accompanied by multiple deformities, and their parents chose to give up at last.
Preoperative and intraoperative acute events
2 cases of preoperative blood oxygen saturation were lower than 90%. There were 35 cases of inhalation pneumonia, 5 of which suffered dyspnea and received endotracheal intubation. The preoperative incidence of acute events in low-birth-weight infants was 66.7% (16/24), and the incidence of acute events in non-low-birth-weight infants was 31.5% (18/57), with statistical difference (P=0.003, table 2). The incidence of preoperative acute events was higher in patients with esophageal atresia and low birth weight. The incidence of intraoperative acute events in patients with low-birth-weight was also higher than that of infants without low-birth-weight (70% vs 26%, P=0.001). The incidence of intraoperative acute events was significantly higher in patients with cardiac malformation (14 cases) than that without cardiac malformation (p=0.001). There was no significant difference in the incidence of preoperative acute events between preterm infants and full-term infants (p=0.06), but the incidence of intraoperative acute events significantly increased in preterm infants (p=0.003).
Postoperative complications in low birth-weight infants
A total of 51 postoperative complications occurred in 81 children (table 2), including 17 postoperative complications in 24 cases of low-birth-weight infants and 34 complications in 57 cases of non-low-birth-weight infants. The postoperative complications in the low-birth-weight infants were significantly higher than that in non-low-birth-weight infants (P=0.029). There were 13 patients suffering postoperative anastomotic fistula and the incidence of anastomotic fistula with low birth weight was higher than non-low-birth-weight infants (p=0.03). Pleural effusion occurred in 5 cases, pneumothorax in 5 cases, and atelectasis in 3 cases. Among the 12 cases of postoperative esophageal stenosis, 10 cases needed esophageal dilatation, and the maximum number of one case was 5 times of esophageal dilatation. Compared with normal-birth-weight infants, the duration of postoperative mechanical ventilation in low-birth-weight infants was 8.5±2.9 days, and that in non-low-birth-weight infants was 3.5±2.5 days, which was significantly prolonged (P=0.001). The duration of ICU monitoring was also prolonged (14.2± 5.2 days vs. 5.9 ±6.1 days, P=0.034), while the thoracic drainage time and hospitalization time were not statistically significant (9.4 ±3.1 days vs. 10.1± 2.2 days, P=0.596; 28± 8.5 days vs. 22.5± 9.1 days, P=0.274, figure 3).
Postoperative stress hyperglycemia
Postoperative stress hyperglycemia in our study refers to the blood glucose concentration (plasma blood glucose, glucose oxidase method) was measured two hours after the postoperative sugar-free rehydration, and the blood glucose concentration higher than 8.0mmol/L is considered to be stress hyperglycemia. A total of 37/81 children with stress hyperglycemia occurred postoperatively, including 13 children with low birth weight (13/24, 54.1%). Postoperative incidence of anastomotic fistula in children with stress hyperglycemia was significantly higher than that in children with normal postoperative blood glucose (P=0.015), which showed postoperative stress hyperglycemia will cause postoperative complications increased. We tested the correlation between postoperative blood glucose level and survival rate, and found a negative correlation (Pearson correlation coefficient -0.931, P=0.022, figure 4).
Following-up
There were ten patients who died after discharge and 27 patients suffered gastroesophageal reflux, which were treated conservatively. 19 children still needed esophageal dilatation after discharge, and 2 children were lower in height and weight than children at the same age. Logistic regression analysis showed that cardiac malformation was the first factor to be removed (P=0.236), while birth weight (P=0.012) and postoperative stress hyperglycemia (P=0.048) were independent prognostic factors.