Introduction: Closed suction drains after pancreaticoduodenectomy (PD) are
routinely used to drain collections and control potential anastomotic leaks. It
is unknown whether the presence of a drain actually affects outcome.
Methods: Eighty-nine patients underwent PD for presumed periampullary
malignancy within the confines of a perioperative nutritional study between
3/1/94 and 8/30/96. Thirty-eight had no intra-abdominal drains and 51 had drains
placed at the conclusion of operation. We analyzed patient, nutritional,
laboratory and operating room factors with endpoints being complications and
length of hospital stay. Intra-abdominal complications were defined as
intra-abdominal abscess and pancreatic or biliary fistula. Analysis was by
Students t-test and chi-square. Significance defined as p<0.05.
Results: Two of 8 surgeons contributed 92% of the patients without drains.
The groups were equivalent with respect to tumor size, comorbid risks and
preoperative nutritional status. There was one perioperative death due to
hepatic artery thrombosis. *mean
No Drains(n=38) Drains(n=51) p
preop tbili (mg/dl)* 3.7 5.1 .40
preop stent (%) 53% 59% .60
ORtime (min)* 292 386 .0001
Blood Loss(ml)* 1097 1083 .92
Fistula (%) 3% 8% .30
Abscess (%) 0% 6% .13
Hospital stay(days)* 12 12 .70
In 2 patients with and 1 patient without drains, CT guided drainage of
collections were necessary. One patient with and no one without drains were
re-operated for bleeding.
Conclusion: Intra-abdominal drainage did not significantly alter the risk of
fistula, abscess, re-operation or the necessity for CT guided intervention after
PD. Routine use of drains after PD may not be indicated, and should be subjected
to a randomized trial.
