No Need for Routine Drainage After Pancreatic Head Resection: The Dual-Center, Randomized, Controlled PANDRA Trial (ISRCTN04937707)

Ann Surg. 2016 Sep;264(3):528-37. doi: 10.1097/SLA.0000000000001859.

Abstract

Objective: This dual-center, randomized, controlled, noninferiority trial aimed to prove that omission of drains does not increase reintervention rates after pancreatic surgery.

Background: There is considerable uncertainty regarding intra-abdominal drainage after pancreatoduodenectomy.

Methods: Patients undergoing pancreatic head resection with pancreaticojejunal anastomosis were randomized to intra-abdominal drainage versus no drainage. Primary endpoint was overall reintervention rate (relaparotomy or radiologic intervention). Secondary endpoints were clinically relevant pancreatic fistula (grade B/C), mortality, morbidity, and hospital stay. The planned sample size was 188 patients per group.

Results: A total of 438 patients were randomized. Forty-three patients (9.8%) were excluded because no pancreatic anastomosis was performed, and 395 patients (202 drain, 193 no-drain) were analyzed. Reintervention rates were not inferior in the no-drain group (drain 21.3%, no-drain 16.6%; P = 0.0004). Overall in-hospital mortality (3.0%) was the same in both groups (drain 3.0%, no-drain 3.1%; P = 0.936). Overall surgical morbidity (41.8%) was comparable (P = 0.741). Clinically relevant pancreatic fistula (grade B/C: drain 11.9%, no-drain 5.7%; P = 0.030) and fistula-associated complications (drain 26.4%; no drain 13.0%; P = 0.0008) were significantly reduced in the no-drain group. Operation time (P = 0.093), postoperative hemorrhage (P = 0.174), intra-abdominal abscess formation (P = 0.199), biliary leakage (P = 0.382), delayed gastric emptying (P = 0.062), burst abdomen (P = 0.480), wound infection (P = 0.758), and hospital stay (P = 0.487) did not show significant differences.

Conclusions: Omission of drains was not inferior to intra-abdominal drainage in terms of postoperative reintervention and superior in terms of clinically relevant pancreatic fistula rate and fistula-associated complications. There is no need for routine prophylactic drainage after pancreatic resection with pancreaticojejunal anastomosis.

Publication types

  • Randomized Controlled Trial

MeSH terms

  • Drainage*
  • Female
  • Humans
  • Length of Stay
  • Male
  • Middle Aged
  • Multivariate Analysis
  • Operative Time
  • Pancreas / surgery*
  • Pancreatic Fistula / etiology
  • Pancreaticoduodenectomy
  • Postoperative Complications
  • Reoperation
  • Risk Factors