Secondary Logo

Journal Logo

Original Articles

Early drain removal after major pancreatectomy reduces postoperative complications: a single-center, randomized, controlled trial

Dai, Menghua MDa,∗; Liu, Qiaofei MDa; Xing, Cheng MDa; Kleeff, Jorg MD, FACSb; Liao, Quan MDa; Guo, Junchao MDa; Han, Xianlin MDa; Xu, Qiang MDa; Wang, Shunda MDa

Author Information
doi: 10.1097/JP9.0000000000000049
  • Open



Major pancreatic reseciton, including both pancreaticoduodenectomy (PD) and distal pancreatectomy (DP), is an effective surgical procedure to treat tumoral and non-tumoral diseases located in duodenum, distal bile duct, and pancreas. Due to the high incidence of postoperative complications, the prophylactic drains after major pancreatectomy are routinely placed, with the aim to reduce postoperative complications by draining fluid collections. Enhanced recovery after surgery (ERAS) protocols have resulted in substantially reduction in complications for several abdominal operations.[1,2] Avoidance of, or early removal of intra-abdominal drains are key elements of ERAS protocols.[3] However, pathogens can also enter the abdominal cavity along drainage tubes as a path, leading to abdominal infections by contaminating aseptic fluid collections. Long-term placement of drain tubes can potentially compress or lacerate vessels, which may lead to bleeding.[4] Therefore, the rationale of routine prophylactic drain placement after pancreatectomy has been challenged.[5,7] In a single-center randomized controlled trial (RCT), closed suction drains failed to reduce mortality and complication rates.[6] The first multicenter RCT had to be stopped early due to the increased mortality from 3% to 12% in the group without drains.[7] Therefore, it seems not rational to abandon prophylactic drains in all patients after pancreatic resection. Removal of drains for selected patients seems more practical and reasonable to reduce complications.[8,9] A non-randomized and single-center study concluded that removal of drains on postoperative day (POD) 4 significantly reduced the incidence of complications after PD.[10,11]

Although, some studies indicated the potential clinical benefits of early drain removal after major pancreatectomy, evidence from RCT is lacking. This single-center RCT was designed to evaluate the safety and efficacy of early removal of prophylactic drains on POD 3, compared to standard removal on or beyond POD 5 in 144 patients with low risk of pancreatic fistula after major pancreatic surgery.

Materials and methods

Study design

This study was approved by the Ethical Committee of Peking Union Medical College Hospital (PUMCH) and was registered at (NCT02230436) on September 3, 2014. Patients who underwent pancreatectomy from June 2014 to July 2018 were recruited into this study. Inclusion criteria were (1) PD with or without pylorus preserving, or DP with or without spleen preserving; (2) age between 18 and 75 years; (3) drain amylase on POD 1 and 3 less than 5000 U/L; (4) drain output within POD 3 less than 300 mL/d. Exclusion criteria were (1) vascular reconstruction using artificial grafts; (2) postoperative bleeding (International Study Group on Pancreatic Surgery [ISGPS] grades B, C), evident anastomosis leakage before randomization within 3 days after surgery. All patients signed the written informed consent before surgery. On POD 3, the patients who met the inclusion criteria were enrolled and they were randomly assigned into 2 groups by central randomization system, generated by the Medical Research & Biometrics Center in National Center for Cardiovascular Disease. Drains were removed on POD 3 in group A (early removal group). In group B (routine removal group), drains were removed on or beyond POD 5 at the discretion of the surgeon. All patients were routinely followed up in outpatient clinic or by telephone interview until the end of the third month after surgery. Demographic data, comorbidities, neoadjuvant/perioperative therapies, surgical procedures, pathological data, medical costs, and postoperative complications were extracted from the electronic medical record system. An electronic standardized case report form was designed to collect the data. The enrollment and randomization flowchart are shown in Figure 1.

Flow diagram of participants in this study. AVD = amylase value in drain; DP = distal pancreatectomy; PD = pancreaticoduodenectomy; POD = postoperative day.

Surgical techniques

Only experienced pancreatic surgeons at the Pancreatic Surgical Center of PUMCH participated in this trial. All surgeons performed more than 30 cases of major pancreatectomy per year and had performed more than 100 cases before recruitment of the first patient. Technical aspects, such as minimally invasive vs open procedures, or classical vs pylorus preserving PD, or closure of the pancreatic remnant by stapler or suture was left to the discretion of the surgeon. In case of DP, a splenectomy was carried out if the tumor was suspected to be malignant or larger than 5 cm in diameter; in other cases, it was left to the discretion of the surgeon. For pancreatic anastomosis, a modified Blumgart duct to mucosa pancreaticojejunostomy was routinely performed. For biliary anastomosis, a single layer running suture was carried out. A 2-layer running gastrojejunostomy or duodenojejunostomy was performed antecolic for all cases. Weather to put bile or pancreatic ductal stent or not when to perform pancreaticojejunostomy and hepaticojejunostomy was determined by the surgeons. Two non-sucking drain tubes were placed underneath the pancreatic and biliary anastomosis or nearby the pancreatic remnant and in the splenic space.

Perioperative managements

All patients were advised to perform pulmonary exercise by using a commercial device (CARENT, Zhejiang, China) before surgery. For obstructive jaundiced patients, pre-operative bile drainage was performed if total bilirubin was greater than 250 μmol/L or acute cholangitis occurred. Restrictive fluid management was adopted (30–35 mL/kg of fluid, per day). Intravenous transfusion of 20 g albumin following by 10 mg furosemide per day were used until POD 3 to reduce edema of the anastomosis and ascites. Somatostatin analogues (intravenous injection of Stilamin, 6 mg/50 mL saline for 3 days and then subcutaneous injection of 0.1 mg Sandostatin, Q8 h, for 2 days) were used within 5 days after surgery. Intravenous Cefmetazole (3.0 g, intravenous injection, twice per day) and Metronidazole (0.915 g, intravenous injection, twice per day) were used to prevent infections within 2 days after surgery. Oral intake of water was restarted on POD 2 and liquid food was restarted on POD 3. The gastric tube was routinely removed on POD 1 or 2. Compressive stockings were used to prevent deep venous thrombosis, and heparin was not routinely used.

Definitions of complications

The Clavien-Dindo classification system was used to grade severity of complications.[12] Intra-abdominal fluid collection was defined as collection measuring at least 3 cm in diameter demonstrated by imaging modalities without signs of infections.[5] Intra-abdominal infections required positive results of germ culture, persistent fever or/and elevated leukocytes.[5] postoperative pancreatic fistula (POPF),[13] postoperative hemorrhage,[14] chyle leak,[15] delayed gastric emptying[16] were defined according to recommendations of the ISGPS.

Statistical analysis

Considering grades 2 to 4 complications could significantly change the clinical management, we adopted the grades 2 to 4 complications as primary end point. The secondary end points were in hospital stay and medical costs. Based on our own and published data, the estimated overall incidence of grades 2 to 4 complications was expected to be 15% of the patients after major pancreatic surgery with low risk of POPF.[5,17,18] This clinical trial was designed to be a superiority study. The clinical relevant target incidence of grades 2 to 4 complications was set at 5% in the early removal of drains group. A total of 144 cases of patients were needed to be compared by the χ2 test, to achieve 80% power in detecting this difference at a 2-sided level of significance of 5%. Participants were allocated 1:1 into group A and group B. Measurement data were presented as mean ± SD. First, the number of patients who had grades 2 to 4 complications were compared (if 1 patient had more than one grades 2–4 complications, the highest grade one was taken as representative); second, the sum of grades 2 to 4 intra-abdominal complications occurrences were compared. Variables between the 2 groups were compared by using χ2 test, Fisher exact test, Wilcoxon rank sum test as appropriate. Logistic regression was adopted for multivariate analysis. A P value less than .05 was considered statistically significant. The SPSS 22.0 software pack (SPSS Inc, Chicago, IL) was used for statistical analysis.


Patient demographics, surgical, and pathological details

A total of 301 patients that underwent major pancreatic surgery were evaluated, 157 cases were excluded. A total of 144 cases were enrolled into the trial and randomly assigned to group A and group B. The demographic profile of the patients in 2 groups was not significantly different (Table 1). Only 1 case in group A, 3 cases in group B underwent neoadjuvant chemotherapy. Pre-operative biliary drainage to relieve obstructive jaundice was performed in 18.1% of the patients in group A and 9.7% of the patients in group B (P = .148). Pathological diagnosis was not statistically different in the 2 groups (Table 1).

Table 1
Table 1:
Demographic characteristic, surgical, and pathologic details of the whole cohort.

Laparoscopic procedures were performed in 40.3% of the patients in group A and 29.2% of the patients in group B (P = .161). Four patients in group A had a soft pancreatic texture vs 7 cases in group B (P = .347). The main pancreatic duct was not dilated (<5 mm) in 50% and 52.8% of the patients in group A and group B (Table 1). The 10-point Fistula Risk Score (FRS) system was calculated in PD patients. Majority of the patients had intermediate risk of POPF and there was no significant difference in the proportion of patients in the low, intermediate and high risk category between the 2 groups. After stratification for PD and DP, the demographic characteristics, comorbidities, pathological, and surgical details were also not significantly different between the 2 groups (all P > .05). A total of 75% and 80% of the DP were performed laparoscopically in group A and group B (Table 2).

Table 2
Table 2:
Demographic characteristic, surgical, and pathologic details of the patients of PD and DP.

Postoperative complications

Only 2 cases of grade B POPF occurred in group A, and no grade B POPF in group B. There was no grade C POPF in any group. Grades 2 to 4 complications occurred in 5 patients in group A (6.9%) as compared to 15 patients in group B (20.8%, P = .028) (Table 3). The sum of grades 2 to 4 intra-abdominal complications in group A was significantly lower than that in group B (11.1% vs 30.6%, P = .026). In detail, the incidence of intra-abdominal infections, delayed gastric emptying, postoperative hemorrhage, biliary fistula, and chylous fistula tended to be lower in group A, while the incidence of pulmonary complications (atelectasis and pleural effusion) seemed to be higher in group A (15.3% vs 8.3%, P = .197) (Table 3). The incidence of intra-abdominal collections was very low in both groups, and was not significantly different. In each group, there were 3 patients that underwent interventional treatment and 1 patient that underwent reoperation. Three patients were readmitted in group A and four patients were readmited in group B. There was no mortality in either group. The in-hospital stay and medical costs were not significantly different between groups A and B (Table 3).

Table 3
Table 3:
Postoperative complications of the whole cohort.

Analysis of the complications according to PD and DP showed that all of the 15 cases of grades 2 to 4 complications in group B and 4 of 5 cases in group A occurred after PD. The incidence of grades 2 to 4 complications in group A was significantly lower than that in group B for PD (P = .0098) but not for DP (P > .99). The sum of grades 2 to 4 intra-abdominal complications in group A was significantly lower than in group B for PD (P = .0095) as well. Intra-abdominal infections in group A seemed to be lower than those in group B for PD, however, this was not significant. One patient underwent reoperation due to incision infection, and the other one due to dislocation of cricoarytenoid joint, both of them were performed under general anesthesia (Table 4).

Table 4
Table 4:
Postoperative complications of the patients of PD and DP.

Risk factors for grades 2 to 4 complications

Univariate analysis showed that the type of operation (PD/DP), laparoscopic vs open procedures, and timing of drain removal (group A vs group B), correlated with grades 2 to 4 complications. PD had a higher incidence of grades 2 to 4 complications as compared to DP (P = .029). Laparoscopic procedures had a lower incidence compared to open procedures (P = .006), and group A had a lower incidence compared to group B (P = .03). After multivariate analysis, only laparoscopic vs open procedures (P = .02, odds ratio = 0.087, 95% CI = 0.011–0.681) and timing of drain removal (P = .039, odds ratio = 0.314, 95% CI = 0.105–0.943) were independent factors associated with a lower incidence of grades 2 to 4 complications (Table 5).

Table 5
Table 5:
Univariate and multivariate regression analysis of risk factors influencing the development of complication (grades 2 to 4).


With the development of the surgical instruments, minimally invasive surgery, multidisciplinary perioperative treatment, and centralization, the morbidity and mortality after major pancreatic surgery has been dramatically reduced.[19–21] The value of prophylactic drain placement for major pancreatic surgery has been increasingly questioned.

Whether to put a drain at all, and when to remove the drain are the 2 key questions.

In 2001, Conlon et al[6] reported routine intraperitoneal drainage was not required after DP and PD, demonstrating that drains did not reduce the need for interventional drainage or surgical exploration. In 2016, Witzigmann et al[22] conducted a dual center, randomized controlled study with 395 cases undergoing pancreatic head resection and concluded that the no-drain approach significantly reduced reintervention rates and clinical-relevant POPF and fistula-associated complications. However, in 2014, Van Buren et al[7] reported that PD without drains was associated with an increase in morality from 3% to 12%. In 2015, another multicenter prospective randomized clinical trial enrolled 137 patients undergoing PD and found that the clinically relevant POPF rate was higher in the no-drain group compared to the drain group, especially for patients with high risk factors for POPF.[23] Therefore, a tailored approach seems currently the most appropriate, and the question is not whether to put a drain, but how to choose the subpopulation of patients who could benefit from the no-drain approach or—most likely safer—early-removal of drains. Molinari et al[11] evaluated the value of amylase value in drain (AVD) on POD 1 to predict POPF and found that a cutoff 5000 U/L of AVD on POD 1 was a stronger predictor than soft pancreas texture. The same group conducted a single-center prospective randomized trial to test the safety of early drain removal (POD 3) after standard pancreatectomy in 114 patients with low risk of POPF as defined by the mentioned cutoff. This study concluded early drain removal reduced the incidence of POPF, abdominal complications and pulmonary complications and decreased hospitalization stay and costs.[5] However, in their study, the incidence of grade B and C POPF was up to 24.6% in the patients with low risk of POPF when drains were removed routinely.[5,10,11] In addition, minimally invasive pancreatic surgery was still in its infant stage at the time of that study, whereas nowadays minimally invasive pancreatic surgery has been well established, particularly for DP.[24–26] Considering of the high incidence of grade B and C POPF in the mentioned study, we restricted the inclusion criteria to further reduce the risk of grade B and C POPF by adding the output of drains within 3 days after surgery and the AVD on POD 3.

In this study, only 2 cases of grade B POPF occurred in 144 enrolled patients, both of which occurred in group A, and no grade C POPF was observed. Early drain removal significantly reduced the incidence of grades 2 to 4 complications from 20.8% to 6.9%. Only 1 case of grade 3 complication occurred after DP and all of the other 19 cases occurred after PD. Early removal of drains significantly reduced the incidence of major complications only in PD but not in DP. Early removal of drains also reduced grades 2 to 4 intra-abdominal complications in PD. Early drain removal was assumed to reduce the incidence of intra-abdominal infections and postoperative bleeding. In this study, the sum incidence of intra-abdominal infections and postoperative bleeding in group A was lower than that in group B (6 vs 13), but this was not significant. Although early drain removal may increase the risk of intra-abdominal collection, there were only 3 cases in group A, compared to 2 cases in group B. In the early drain removal group, there was also a decreased trend for biliary fistula, chyle fistula, and delayed gastric emptying. In the aforementioned study,[5] early drain removal significantly reduced abdominal complications from 52.6% to 12.2% and decreased the incidence of pulmonary complication. However, in this study, we found that early drain removal had a tendency to increase the incidence of pulmonary complications. It has also been shown that early drain removal reduces postoperative stay and costs[7]; however, in the present study, postoperative stay and costs were not significantly different between groups. The reasons for these results may be due to the low overall incidence of grade B and C POPF and grades 2 to 4 complications in our study as well as differences in health care systems and reimbursement. In the present study, the postoperative stay of these 2 groups was not significantly different. There was only case of reoperation in each group under general anesthesia; however, none of them suffered organ dysfunction. The readmission rate in each group was also comparable.

Laparoscopic DP was performed in 75% of the patients in group A and 80% of the patients in group B, and 26.9% and 9.6% of PD were performed laparoscopically in group A and group B, respectively. Laparoscopic procedures significantly reduced the incidence of grades 2 to 4 complications, compared to open procedures. Similar to other studies, a small pancreatic duct and soft pancreatic texture did not correlate with major complications.[5,22,27,28] In this study, only 3 out of 144 patients had high risk of POPF defined by the 10-point FRS system, and the other patients had intermediate risk. These results showed that our criteria were consistent with the 10-point FRS system.

Although this is a RCT, there are some limitations: in this study, both DP and PD were enrolled, however, the surgical complexity and the incidence and severity of complications of PD and DP are different, therefore, it could be better to enroll DP or PD separately with a larger sample size to achieve more strong statistical power; this is only a single-center study, a further multicenter RCT are needed to confirm the results.


Using our criteria, the incidence of grade B or C POPF is extremely low in patients undergoing DP and PD. Early removal of drains on POD 3 is safe for both DP and PD patients. Early drain removal does not affect the complications in DP patients; however, it significantly reduces the overall incidence of grades 2 to 4 complications in patients undergoing PD. Since around half of the patients undergoing major pancreatectomy met those criteria, this study presents high-level evidence to change the routine management of prophylactic drains in selected patients under our criteria after major pancreatectomy.


We thank all of the staff from Department of General Surgery, Peking Union Medical College Hospital, Peking Union Medical College & Chinese Academy of Medical Sciences of China.

Author contributions

All authors have read and approved the submitted manuscript, and it has not been submitted or published elsewhere.

Financial support

This study was supported by the project of application and promotion of capital special clinical research from Beijing Municipal Science & Technology Commission (Z171100001017017018).

Conflicts of interest

The authors declare no conflicts of interest.

Ethics approval

This study was approved by the Ethical Committee of Peking Union Medical College Hospital (PUMCH) and was registered at (NCT02230436).


[1]. Gustafsson UO, Hausel J, Thorell A, et al. Adherence to the enhanced recovery after surgery protocol and outcomes after colorectal cancer surgery. Arch Surg 2011;146:571–577.
[2]. Coolsen MM, Wong-Lun-Hing EM, van Dam RM, et al. A systematic review of outcomes in patients undergoing liver surgery in an enhanced recovery after surgery pathways. HPB (Oxford) 2013;15:245–251.
[3]. Ljungqvist O, Scott M, Fearon KC. Enhanced recovery after surgery: a review. JAMA Surg 2017;152:292–298.
[4]. Kagedan DJ, Ahmed M, Devitt KS, et al. Enhanced recovery after pancreatic surgery: a systematic review of the evidence. HPB (Oxford) 2015;17:11–16.
[5]. Bassi C, Molinari E, Malleo G, et al. Early versus late drain removal after standard pancreatic resections: results of a prospective randomized trial. Ann Surg 2010;252:207–214.
[6]. Conlon KC, Labow D, Leung D, et al. Prospective randomized clinical trial of the value of intraperitoneal drainage after pancreatic resection. Ann Surg 2001;234:493–494. discussion 493–494.
[7]. Van Buren G 2nd, Bloomston M, Hughes SJ, et al. A randomized prospective multicenter trial of pancreaticoduodenectomy with and without routine intraperitoneal drainage. Ann Surg 2014;259:605–612.
[8]. Correa-Gallego C, Brennan MF, D’Angelica M, et al. Operative drainage following pancreatic resection: analysis of 1122 patients resected over 5 years at a single institution. Ann Surg 2013;258:1051–1058.
[9]. Nitsche U, Müller TC, Späth C, et al. The evidence based dilemma of intraperitoneal drainage for pancreatic resection—a systematic review and meta-analysis. BMC Surg 2014;14:76.
[10]. Kawai M, Tani M, Terasawa H, et al. Early removal of prophylactic drains reduces the risk of intra-abdominal infections in patients with pancreatic head resection: prospective study for 104 consecutive patients. Ann Surg 2006;244:1–7.
[11]. Molinari E, Bassi C, Salvia R, et al. Amylase value in drains after pancreatic resection as predictive factor of postoperative pancreatic fistula: results of a prospective study in 137 patients. Ann Surg 2007;246:281–287.
[12]. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–213.
[13]. Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery 2017;161:584–591.
[14]. Wente MN, Veit JA, Bassi C, et al. Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery 2007;142:20–25.
[15]. Besselink MG, van Rijssen LB, Bassi C, et al. Definition and classification of chyle leak after pancreatic operation: a consensus statement by the International Study Group on Pancreatic Surgery. Surgery 2017;161:365–372.
[16]. Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 2007;142:761–768.
[17]. Braga M, Capretti G, Pecorelli N, et al. A prognostic score to predict major complications after pancreaticoduodenectomy. Ann Surg 2011;254:702–707.
[18]. Uzunoglu FG, Reeh M, Vettorazzi E, et al. Preoperative Pancreatic Resection (PREPARE) score: a prospective multicenter-based morbidity risk score. Ann Surg 2014;260:857–863.
[19]. Fogel EL, Shahda S, Sandrasegaran K, et al. A multidisciplinary approach to pancreas cancer in 2016: a review. Am J Gastroenterol 2017;112:537–554.
[20]. Dokmak S, Fteriche FS, Aussilhou B, et al. The largest European single-center experience: 300 laparoscopic pancreatic resections. J Am Coll Surg 2017;225:226–234.e2.
[21]. Griffin JF, Poruk KE, Wolfgang CL. Pancreatic cancer surgery: past, present, and future. Chin J Cancer Res 2015;27:332–348.
[22]. Witzigmann H, Diener MK, Kienkotter S, et al. No need for routine drainage after pancreatic head resection: the dual-center, randomized, controlled PANDRA trial (ISRCTN04937707). Ann Surg 2016;264:528–537.
[23]. McMillan MT, Fisher WE, Van Buren G 2nd, et al. The value of drains as a fistula mitigation strategy for pancreatoduodenectomy: something for everyone? Results of a randomized prospective multi-institutional study. J Gastrointest Surg 2015;19:21–30. discussion 30–31.
[24]. Wright GP, Zureikat AH. Development of minimally invasive pancreatic surgery: an evidence-based systematic review of laparoscopic versus robotic approaches. J Gastrointest Surg 2016;20:1658–1665.
[25]. de Rooij T, van Hilst J, van Santvoort H, et al. Minimally invasive versus open distal pancreatectomy (LEOPARD): a multicenter patient-blinded randomized controlled trial. Ann Surg 2019;269:2–9.
[26]. van Hilst J, de Rooij T, Bosscha K. Laparoscopic versus open pancreatoduodenectomy for pancreatic or periampullary tumours (LEOPARD-2): a multicentre, patient-blinded, randomised controlled phase 2/3 trial. Lancet Gastroenterol Hepatol 2019;4:199–207.
[27]. Ven Fong Z, Correa-Gallego C, Ferrone CR, et al. Early drain removal—the middle ground between the drain versus no drain debate in patients undergoing pancreaticoduodenectomy: a prospective validation study. Ann Surg 2015;262:378–383.
[28]. Callery MP, Pratt WB, Kent TS, et al. A prospectively validated clinical risk score accurately predicts pancreatic fistula after pancreatoduodenectomy. J Am Coll Surg 2013;216:1–14.

Complications, Distal pancreatectomy, Drain removal, Pancreaticoduodenectomy

Copyright © 2020 The Chinese Medical Association, Published by Wolters Kluwer Health, Inc. under the CCBY-NC-ND license.