Clinical effect and safety of nutritional support after pancreaticoduodenectomy: A Meta-analysis

Objective To investigate the effect of different nutritional support methods (including total parenteral nutrition, early oral nutrition, early enteral nutrition via nasojejunal tube, early enteral nutrition via gastric fistula, and early enteral nutrition via jejunal fistula) on the prognosis of patients after pancreaticoduodenectomy (PD) . Methods Pub Med, Embase, and The Cochrane Library were searched for articles on nutritional support after PD published from 2007 to 2017. A pooled analysis was performed for related data including sample size, research contents, study population, research roadmap and methods, length of hospital stay, and incidence rates of infection (including postoperative pulmonary infection and incision infection) , pancreatic fistula (grade B/C) , and delayed gastric emptying, and Rev Man 5. 3 software was used for the meta-analysis of total parenteral nutrition and early enteral nutrition via nasojejunal tube after PD. Results A total of 11 articles were included, and 4 were included in the meta-analysis. The horizontal comparison showed no significant difference in the length of hospital stay between the patients undergoing nutritional support via different pathways, while the vertical comparison showed a significant difference. As for complications, the gastric fistula group had a significantly higher incidence rate of pancreatic fistula than the other groups (all P < 0. 01) , and the jejunal fistula group had a significantly higher incidence rate of delayed gastric emptying than the other groups (all P < 0. 01) . The total parenteral nutrition group had an incidence rate of infection of 27. 17%, which was lower than that in the other groups (40. 63% 、46. 58% 、33. 64%) , but there was no statistical significance in the differences (all P > 0. 05) . The meta-analysis showed no significant differences between the total parenteral nutrition group and the nasojejunal tube early enteral nutrition group in length of hospital stay, delayed gastric emptying, and incidence rates of pancreatic fistula and infection. Conclusion Nutritional strategies after PD have similar clinical effects and safety, and a reasonable nutritional support regimen can be developed based on clinical practice and patients' conditions.