Abstract: Background：Human serum albumin infusion was empirical therapy after neonatal gastric perforation surgery. It is unclear whether excessive supplementation is beneficial. Objective：To investigate whether prolonged human serum albumin infusion treatment in newborns with gastric perforation after abdominal surgery can improve the short-term outcomes. Design：Retrospective cohort study. Methods：The newborns who received intravenous infusion of albumin(1-2 g·kg-1) after gastric perforation were divided into short course group (≤5 days) and long course group (>5 days) according to the time of empirical application of albumin. The perinatal condition, main symptoms, signs, and laboratory results before operation, preoperative and postoperative complications, surgical condition (perforation site, pathological findings, etc.), postoperative treatment and hospital prognosis were collected. Main outcome measures：Postoperative nutritional status and hospitalization outcomes. Results：There were 37 cases in the short course group and 26 cases in the long course group. There was no significant difference in perinatal general condition, preoperative symptoms and signs, preoperative laboratory results, complications, perforation location and pathological findings between the two groups（P＞0.005）. In the short course group compared with the long-term treatment group, the gestational age was younger (P=0.005), the proportion of diabetes in pregnancy was higher(P=0.034), the age of onset was earlier (P=0.035), the age of operation was younger (P=0.035), the total amount of albumin infusion after operation was less (P=0.000), the 70% total enteral nutrition time was longer (P=0.019), the total enteral nutrition time was longer (P=0.005), and the hospital stay was shorter (P=0.012). However, there was no significant difference in secondary operation, length of stay, status of improvement or curation and weight at discharge between the two groups. Neonates with the difference of birth weight < 250 g, gestational age < 1 week and age of onset < 1 d in short course group and long course group were matched at 1:1. Twenty-two cases were included in each group. There was no significant difference in postoperative antibiotic use time, postoperative invasive respiratory support, times of plasma transfusion, gastrointestinal decompression time, fistulostomy tube indwelling time, postoperative enteral feeding time, 70% total enteral nutrition time, total enteral nutrition time, secondary operation, hospital stay and discharge weight between the two groups. Conclusion：Prolonged treatment time of human serum albumin infusion is not beneficial to the nutritional status and hospitalization outcomes of newborns with gastric perforation.

Key words: Gastric Perforation, Human serum albumin, Prognosis, Newborn