Duffas 2005.

Methods	Randomized controlled trial Multicenter
Participants	Country: France. Number randomized: 149. Post‐randomization dropout: 0 (0%). Mean age: 58.4 years. Female: 63 (42.3%). Pancreatic cancer: 59 (39.6%). Biliary cancer: 19 (12.8%). Ampullary cancer: 36 (24.1%). Duodenal cancer: 6 (4.0%). Other: 29 (19.5%). Classic pancreaticoduodenectomy: 113 (75.8%). Pylorus‐preserving pancreaticoduodenectomy: 36 (24.2%). Inclusion criteria: Participants aged ≥ 18 years. Participants with pancreatic (malignant or benign) tumor or chronic pancreatitis or extrapancreatic tumor (ampullar, biliary, or duodenal). Exclusion criteria: Participants aged < 18 years. Participants underwent bilio‐ or gastroenteric bypass, cystic or pancreatic drainage without resection. Simple tumor excision. Resection for acute pancreatitis or trauma. Total pancreatectomy. Pancreaticoduodenectomy without immediate pancreatic anastomosis or duodenum‐preserving pancreatectomy.
Interventions	Participants (N = 149) were randomly assigned to two groups. Group 1: Pancreaticojejunostomy (N = 68). Group 2: Pancreaticogastrostomy (N = 81).
Outcomes	Postoperative pancreatic fistula: Yes. Postoperative mortality: Yes. Length of hospital stay: Yes. Rate of surgical re‐intervention: Yes. Overall rate of surgical complications: Yes. Rate of postoperative bleeding: Yes. Rate of intra‐abdominal abscess: Yes. Quality of life: No. Cost analysis: No.
Notes	Definition of pancreatic fistula: (1) chemically as fluid obtained through drains or percutaneous aspiration, containing at least 4 times normal serum values of amylase for 3 days, irrespective of the amount of output and the date of appearance or (2) clinically and radiologically, as anastomotic leaks shown by fistulography. The calculated sample size, based on the reduction of intra‐abdominal complications rate from 40% to 20%, was 134 participants and the study met its estimated accrual.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "computerized random number tables".
Allocation concealment (selection bias)	Low risk	Quote: "Random allotment was through a telephone call to the coordinating center".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Postoperative complications were assessed by a physician who was unaware of the allotted treatment".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "There were no protocol violations, no crossovers, or withdrawals after randomization". Comment: There were no post‐randomization dropouts.
Selective reporting (reporting bias)	Low risk	Comment: All the primary outcomes were reported. The review authors consider this study to be free of selective reporting for the primary outcomes.
Other bias	Low risk	Comment: The study appears to be free of other sources of bias.