Yes, placing drains is essential.

It’s important to look at the evidence in the literature, starting with a randomized controlled trial of 179 patients from 2001 that showed no reduction in death or complication rate associated with use of surgical intraperitoneal closed suction drainage (Ann. Surg. 2001;234:487-94). Interestingly, even though this study showed there was no benefit to using drains, they address the use of closed suction drainage. The investigators were not claiming all drains are unnecessary.

Why drain placement? The argument for drains includes evacuation of blood, pancreatic juice, bile, and chyle. In addition, assessing drainage can act as a warning sign for anastomotic leak or hemorrhage, so patients can potentially avoid additional interventions.

However, the clinically relevant question with drains during pancreatic resection is, Do drains increase the rates of fistula or infection?

A study from University of Tokyo researchers found three drains were more effective than one. In addition, the investigators argued against early removal, pointing out that the risk for infection associated with drains only increased after day 10. The researchers were not only in favor of drains but in favor of multiple drains (World J Surg. 2016;40:1226-35).

A multicenter, randomized prospective trial conducted in the United States compared 68 patients with drains to 69 others without during pancreaticoduodenectomy. They reported an increase in the frequency and severity of complications when drains were omitted, including the number of grade 2 or greater complications. Furthermore, the safety monitoring board stopped the study early because mortality among patients in the drain group was 3%, compared with 12% in the no-drain group. (Ann. Surg. 2014:259:605-12).

Another set of researchers in Germany conducted a prospective, randomized study that favored omission of drains. However, a closer look at demographics shows that about one-fourth of participants had chronic pancreatitis, which is associated with a low risk of fistula (Ann Surg. 2016;263:440-9). In addition, they found no significant difference in fistula rates between patients who underwent pancreaticojejunostomy or pancreaticoduodenectomy. Interestingly, the authors noted that surgeons were reluctant to omit drains in many situations, even in a clinical trial context.

Furthermore, a systematic review of nine studies with nearly 3,000 patients suggests it is still necessary to place abdominal drains during pancreatic resection, researchers at the Medical College of Xi’an Jiaotong University in China reported (World J Gastroenterol. 2015;21:5719-34). The authors cited a significant increase in morbidity among patients in whom drains were omitted (odds ratio, 2.39).

Perhaps the cleverest way to address this controversy with drains during pancreatic resection comes from researchers at the University of Pennsylvania. They looked at surgical risk factors to gain a more nuanced view. They used the Fistula Risk Score (FRS) to stratify patients and re-examined the outcomes of the multicenter U.S. study I cited earlier that was stopped early because of differences in mortality rates. The University of Pennsylvania research found that FRS correlated well with outcomes, suggesting its use as a mitigation strategy in patients at moderate to high risk for developing clinically relevant postoperative pancreatic fistula (J Gastrointest Surg. 2015;19:21-30). In other words, they suggest routine prophylactic drainage for patients at moderate to high risk but also suggest that there may be no need for prophylactic drainage for negligible to low risk patients.

I would like to conclude that drains appear essential in many cases, including for moderate to higher risk patients. We need to tailor our practices as surgeons regarding placement of drains during pancreatectomy or pancreaticoduodenectomy. Our goal remains to create opportunities for prolonged survival and better quality of life for our patients.

Dr. Dervenis is head of the Department of Surgical Oncology at the Metropolitan Hospital in Athens, Greece. He is also chair of the hepato-pancreato-biliary surgical unit. He reported no disclosures.

No, drain placement is not always necessary.

Drain placement is not essential during all pancreatectomies. The practice is so historically entrenched in our specialty that I need you to check your dogma at the door and take a look at the data with an open mind.

There are many, many retrospective studies that examine drainage versus no drainage. All of these studies have shown the same thing: There is no difference in many outcomes, including mortality. There may be a difference in terms of complicated, postoperative fistulae, which are difficult to manage in patients who have drains.

Researchers at Memorial Sloan-Kettering Cancer Center (MSKCC) in New York City conducted the first prospective trial addressing this issue (Ann. Surg. 2001;234:487-94). They evaluated 179 patients, randomizing 88 patients to receive intraoperative drains. Patients in the cohort with drains were more likely to develop a significant intra-abdominal abscess, collection, or fistula.

Another prospective randomized trial from Germany compared the reintervention rate, an endpoint I like, among 439 patients (Ann Surg. 2016;264:528-37). There was no statistically significant difference between the drain or no drain groups, further suggesting that drains are not essential in all cases.

We also need to take a closer look at the study Dr. Dervenis cited (Ann Surg. 2014:259:605-12). It’s true this trial was stopped early because of a higher mortality rate of 12% in the no drain group, but the difference was not statistically significant (P = .097). There was no difference in the fistula rate.

When you look at these data, there are a couple of conclusions you can reach. One conclusion is that a trial of 130 people designed for 750 found something uniquely different that had never been reported in any retrospective series, including one by the study’s senior author that demonstrated drains are essential (HPB [Oxford]. 2011;13:503-10). Another conclusion is that drains are not needed, and there is a very good chance this is a false positive finding. Look at the reasons the 10 participants died. I’m not sure a drain would have helped some of the patients without them, and some of the patients who had drains still died.

A study looking at practice at my institution, MSKCC, shows that drains are still used about half the time, indicating they are not essential. (Ann. Surg. 2013 Dec;258[6]:1051-8). Drains were more commonly used for pancreaticoduodenectomy and when the surgeon thought there might be a problem, such as a soft gland, difficult time in the OR, or a small pancreatic duct.

If you look at these data, this pans out in every retrospective study I’ve seen comparing drain versus no drain – the patients with drains tend to have higher morbidity. Among the 1,122 resection patients in the MSKCC study, those without operative drains had significantly lower rates of grade 3 complications and overall morbidity and fewer readmissions and lower rates of grade 3 or higher pancreatic fistula. Mortality and reintervention rates were no different.

I also looked at our most recent data, between 2010 and 2015. Now we are using intraoperative drains even less often. They are not considered essential at this point.

There are multiple retrospective and well-designed prospective, randomized trials that show no benefit to routine drainage following pancreatectomy. Acceptance of randomized clinical data is slow, particularly when it flies in the face of what you were taught by your mentors over the past 40 or 50 years. I encourage you to look at these data carefully and utilize them in your practice.

Dr. Allen is associate director for clinical programs at David M. Rubenstein Center for Pancreatic Cancer Research and the Murray F. Brennan chair in surgery at Memorial Sloan-Kettering Cancer Center in New York City. He reported no disclosures.

Point/Counterpoint: Is intraoperative drain placement essential during pancreatectomy?