Sarr 2003.

Methods	Randomised clinical trial (parallel design)
Participants	Country: USA Sample size: 381 Post‐randomisation drop‐out: 106 Revised sample size: 275 Females: 95 (34.5%) Mean age: 62 years Malignancy: 138 (50.2%) Chronic pancreatitis: 0 (0%) Pancreatoduodenectomy: 108 (39.3%) Follow‐up: 30 days Inclusion criteria: 1. People undergoing an elective, anatomic, proximal or central pancreatectomy with a pancreaticoenteric anastomosis or a distal pancreatectomy with or without a pancreaticoenteric anastomosis 2. At least 18 years of age 3. An elective pancreatic resection because of a presumed pancreatic or periampullary neoplasm Exclusion criteria: 1. Chronic pancreatitis 2. Emergency surgery 3. Total pancreatectomy 4. Duct‐drainage procedures alone without any pancreatic resection 5. Enucleation of neoplasm 6. Enteric drainage of a pancreatic pseudocyst 7. Serum creatinine of twice the local upper limit of normal 8. Serum bilirubin > 20 mg/dL 9. Administration of a somatostatin analogue within the previous month
Interventions	The participants were randomly assigned to 2 groups Group 1: vapreotide (n = 135) Further details: 0.6 mg sc 2 h before operation and then twice daily for 7 days postoperatively Group 2: placebo (n = 140) Further details: same times as intervention
Outcomes	The outcomes reported were mortality and perioperative morbidity
Notes	Reasons for post‐randomisation drop‐out: did not undergo pancreatic resection Pancreatic fistula definition: drainage fluid (beginning on or after postoperative day 5) of both > 30 mL/day and amylase or lipase activity > 5‐fold upper limits of the normal serum value (grade A)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: this information was not available
Allocation concealment (selection bias)	Unclear risk	Comment: this information was not available
Blinding (performance bias and detection bias) All outcomes	Low risk	Quote: "patients received vapreotide acetate (0.6 mg) or placebo subcutaneously within up to 2 hours before operation and then twice daily for 7 days postoperatively"
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "patients not resected were replaced" Comment: this would have necessitated exclusion of the trial if there was no blinding (as replacing recruited patients with other patients introduces selection bias). However, we have treated the patients who were replaced as drop‐outs because of adequate blinding
Selective reporting (reporting bias)	High risk	Comment: all pre‐defined outcomes were not reported
Free of source of funding bias?	Low risk	Quote: "we would like to thank J‐M Dumont and K Besseghir from Debiopharm SA for their guidance and financial support" Comment: in spite of obtaining financial support from a pharmaceutical company, the authors of this report do not support the use of the drug