Prophylactic abdominal drainage for pancreatic surgery
Abstract
Background
The use of surgical drains has been considered mandatory after pancreatic surgery. The role of prophylactic abdominal drainage to reduce postoperative complications after pancreatic surgery is controversial.
Objectives
To assess the benefits and harms of routine abdominal drainage after pancreatic surgery, compare the effects of different types of surgical drains, and evaluate the optimal time for drain removal.
Search methods
We searched The Cochrane Library (2015, Issue 3), MEDLINE (1946 to 9 April 2015), EMBASE (1980 to 9 April 2015), Science Citation Index Expanded (1900 to 9 April 2015), and Chinese Biomedical Literature Database (CBM) (1978 to 9 April 2015).
Selection criteria
We included all randomized controlled trials that compared abdominal drainage versus no drainage in patients undergoing pancreatic surgery. We also included randomized controlled trials that compared different types of drains and different schedules for drain removal in patients undergoing pancreatic surgery.
Data collection and analysis
Two review authors independently identified the trials for inclusion, collected the data, and assessed the risk of bias. We performed the meta‐analyses using Review Manager 5. We calculated the risk ratio (RR) for dichotomous outcomes and the mean difference (MD) for continuous outcomes with 95% confidence intervals (CI). For all analyses, we employed the random‐effects model.
Main results
Drain use versus no drain use
We included two trials involving 316 participants who were randomized to the drainage group (N = 156) and the no drainage group (N = 160) after pancreatic surgery. Both trials were at high risk of bias. There was inadequate evidence to establish the effect of drains on mortality at 30 days (drains 1.3%; no drains 3.8%; RR 0.44; 95% CI 0.05 to 3.94; two studies; very low‐quality evidence), mortality at 90 days (2.9% versus 11.6%; RR 0.24; 95% CI 0.05 to 1.10; one study; very low‐quality evidence), intra‐abdominal infection (8.3% versus 14.4%; RR 0.61; 95% CI 0.25 to 1.46; two studies), wound infection (10.9% versus 11.9%; RR 0.91; 95% CI 0.45 to 1.86; two studies), morbidity (67.3% versus 65.0%; RR 1.02; 95% CI 0.88 to 1.19; two studies), length of hospital stay (MD ‐0.97 days; 95% CI ‐1.41 to ‐0.53; two studies), or additional open procedures for postoperative complications (6.3% versus 6.4%; RR 0.90, 95% CI 0.15 to 5.32; two studies). There was one drain‐related complication in the drainage group (0.6%). The quality of evidence was low, or very low.
Type of drain
There were no randomized controlled trials comparing one type of drain versus another.
Early versus late drain removal
We included one trial involving 114 participants with a low risk of postoperative pancreatic fistula who were randomized to the early drain removal group (N = 57) and the late drain removal group (N = 57) after pancreatic surgery. The trial was at high risk of bias. There was no evidence of differences between the two groups in the mortality at 30 days (0% for both groups) or additional open procedures for postoperative complications (0% versus 1.8%; RR 0.33; 95% CI 0.01 to 8.01). The early drain removal group was associated with lower rates of postoperative complications (38.5% versus 61.4%; RR 0.63; 95% CI 0.43 to 0.93), shorter length of hospital stay (MD ‐2.10 days; 95% CI ‐4.17 to ‐0.03; 21.5% decrease of an 'average' length of hospital stay) and hospital costs (17.0% decrease of 'average' hospital costs) than in the late drain removal group. The quality of evidence for each of the outcomes was low.
Authors' conclusions
It is not clear whether routine abdominal drainage has any effect on the reduction of mortality and postoperative complications after pancreatic surgery. In case of drain insertion, low‐quality evidence suggests that early removal may be superior to late removal for patients with low risk of postoperative pancreatic fistula.
PICOs
Plain language summary
Drain use after pancreatic surgery
Review question
Is drain use able to reduce postoperative complications after pancreatic surgery?
Background
The use of surgical drains has been considered mandatory after pancreatic surgery. The role of drain use to reduce postoperative complications after pancreatic surgery is controversial.
Study characteristics
We searched for all relevant, well‐conducted studies up to April 2015. We included three randomized controlled trials (an experiment in which participants are randomly allocated to two or more interventions, possibly including a control intervention or no intervention, and the results are compared). A total of 430 participants were included in the three studies. Two of the three trials randomized a total of 316 patients to drain use (N = 156) or no drain use (N = 160). The other trial randomized a total of 114 patients with low risk of postoperative pancreatic fistula (a complication during which the pancreas is disconnected from the nearby gut, and then reconnected to allow pancreatic juice containing digestive enzymes to enter the digestive system) to early drain removal (N = 57) and late drain removal (N = 57).
Key results
There was insufficient evidence to determine the effect on death (1.3% versus 3.8%), intra‐abdominal infections (8.3% versus 14.4%), wound infections (10.9% versus 11.9%), overall complications (67.3% versus 65.0%), duration of hospitalization (8.1 days versus 8.6 days), or additional open procedures for postoperative complications (6.3% versus 6.4%) between drain use and no drain use. There was one drain‐related complication (the drainage tube was broken) in the drain use group (0.6%). Information on deaths in early or late removal of drains was available from one small study in which there were no deaths reported in either group. Rates of additional open procedures for postoperative complications were low (0% in early removal versus 1.8% in late). Early drain removal was associated with a lower complication rate (38.5% versus 61.4%), shorter length of hospital stay (21.5% decrease of an 'average' length of hospital stay) and lower hospital costs (17.0% decrease of 'average' hospital costs) than in the late drain removal.
It is not clear whether routine drain use has any effect on the reduction of postoperative complications after open pancreatic surgery. In the case of drain insertion, early removal appears to be better than late removal for patients with a low risk of postoperative pancreatic fistula.
Quality of the evidence
All trials were at high risk of bias (suggesting the possibility of overestimating the benefits or underestimating the harms). Overall, the quality of the evidence varied from very low to low.
Authors' conclusions
Summary of findings
| Drain use versus no drain use for pancreatic surgery | ||||||
| Patient or population: Patients undergoing elective open pancreatic resections | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| No drain use | Drain use | |||||
| Mortality (30 days) | Study population | RR 0.44 | 316 | ⊕⊝⊝⊝ | ||
| 38 per 1000 | 16 per 1000 | |||||
| Moderate | ||||||
| 40 per 1000 | 18 per 1000 | |||||
| Mortality (90 days) | Study population | RR 0.24 | 134 | ⊕⊝⊝⊝ | ||
| 121 per 1000 | 29 per 1000 | |||||
| Moderate | ||||||
| 121 per 1000 | 29 per 1000 | |||||
| Intra‐abdominal infection | Study population | RR 0.61 | 316 | ⊕⊝⊝⊝ | ||
| 144 per 1000 | 88 per 1000 | |||||
| Moderate | ||||||
| 156 per 1000 | 95 per 1000 | |||||
| Drain‐related complications | Study population | RR 3.1 | 179 | ⊕⊝⊝⊝ | There was one drain‐related complication in the drainage group. The drainage tube was broken. | |
| 0 per 1000 | 12 per 1000 | |||||
| Moderate | ||||||
| 0 per 1000 | 0 per 1000 | |||||
| Morbidity | Study population | RR 1.02 | 316 | ⊕⊝⊝⊝ | ||
| 650 per 1000 | 663 per 1000 | |||||
| Moderate | ||||||
| 663 per 1000 | 676 per 1000 | |||||
| Length of hospital stay | The mean length of hospital stay in the no drain group was 8.12 days | The mean length of hospital stay in the drain groups was | MD ‐0.97 (‐1.41 to ‐0.53) | 316 | ⊕⊝⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level for serious risk of bias. 5 Downgraded one level for serious imprecision (Total population size was less than 400). | ||||||
| Early versus late drain removal for pancreatic surgery | ||||||
| Patient or population: Patients undergoing elective open pancreatic resections | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Late drain removal | Early drain removal | |||||
| Mortality (30 days) | See comment | See comment | Not estimable | 114 | ⊕⊕⊝⊝ | There was no mortality in either group. |
| Morbidity | Study population | RR 0.63 | 114 | ⊕⊕⊝⊝ | ||
| 614 per 1000 | 387 per 1000 | |||||
| Moderate | ||||||
| 614 per 1000 | 387 per 1000 | |||||
| Length of hospital stay (days) | The mean length of hospital stay in the late removal group was 10.8 days | The mean length of hospital stay in the early removal group was | MD ‐2.10 (‐4.17 to ‐0.03) | 114 | ⊕⊕⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level due to serious risk of bias. | ||||||
Background
Description of the condition
Pancreatic cancer ranks thirteenth in terms of the most common cancers and eighth as the cause of cancer death from a global viewpoint (Anderson 2006; Lowenfels 2006). Regional differences exist in the incidence, the number of new patients diagnosed per year (Anderson 2006). The overall incidence of pancreatic cancer is approximately 4 to 10 cases per 100,000 persons per year (Dragovich 2011). The most common cause of pancreatic cancer is heavy tobacco usage (Lowenfels 2006).
Although the exact incidence of chronic pancreatitis (long‐standing inflammation of the pancreas) worldwide is unknown, the estimated incidence of chronic pancreatitis is six cases per 100,000 persons per year in Europe and probably all western countries (Spanier 2008). The prevalence (the total number of patients at a designated time) of chronic pancreatitis in the United Kingdom, France, Japan, and south India is 3 cases, 26 cases, 4 cases, and 114 to 200 cases per 100,000 persons, respectively (Bornman 2001; Braganza 2011; Garg 2004; Lévy 2006). The most common cause of chronic pancreatitis is alcohol abuse (Braganza 2011; Spanier 2008).
Pancreatic surgery is performed to treat various pancreatic and extra‐pancreatic diseases, including pancreatic cancers, chronic pancreatitis, biliary (related to the bile duct) and duodenal (related to the first section of the small intestine) malignancy, etc (Cheng 2013a; Connor 2005; Gurusamy 2013). Although the mortality (the proportion of deaths after surgery) of pancreatic surgery has been reduced to less than 5% currently, the overall morbidity (the proportion of patients with any postoperative complications) is still high, ranging from 30% to 60% (Bassi 2005; Connor 2005; Giovinazzo 2011; Gurusamy 2013; Wente 2007a; Wente 2007b). The most common complications after pancreatic surgery include delayed gastric emptying (19% to 23%) (Wente 2007a; Wente 2007b), pancreatic fistula (2% to 30%; a complication whereby the pancreas is disconnected from the nearby gut, and then reconnected to allow pancreatic juice containing digestive enzymes to enter the digestive system; Bassi 2005; Cheng 2012a; Hackert 2011; Wente 2007a; Wente 2007b), intra‐abdominal abscess (9% to 10%; Wente 2007a; Wente 2007b), wound infection (5% to 15%; Andrén‐Sandberg 2011; Halloran 2002), and postoperative bleeding (1% to 8%; Wente 2007a; Wente 2007b).
Description of the intervention
As a measure to reduce postoperative complications, prophylactic drains are traditionally placed in the subhepatic space near both the biliary and pancreatic anastomoses (the surgical connection of the bile‐pancreatic duct and gut to form a continuous channel) after pancreatic resections (Conlon 2001; Fisher 2011). Abdominal drainage has been in use for over 1000 years (Memon 2001).
Surgical drains are artificial tubes used to remove blood, pus, or other body fluids from wounds (Durai 2009). There are two main types of surgical drains: open and closed (Cheng 2012b; Durai 2009; Gurusamy 2007a; Wang 2011). An open drain communicates with the atmosphere (e.g., corrugated drain, Penrose drain, sump drain; Durai 2009; Gurusamy 2007a; Wang 2011). A closed drain consists of a tube that drains into a collection bag or bottle where the contents are not exposed to the atmosphere (Durai 2009; Gurusamy 2007a; Wang 2011). Closed drains may be either active (suction drains under low or high pressure, e.g., Jackson‐Pratt drain, Redivac) or passive (drains without suction, e.g., Robinson drain, Pigtail drain; Durai 2009; Gurusamy 2007a; Wang 2011).
How the intervention might work
Surgeons have routinely used drains after pancreatic surgery because of the fear of collection of bile, pancreatic juice, or blood, which may require additional procedures (Adham 2013; Bassi 2010; Conlon 2001; Correa‐Gallego 2013; Fisher 2011; Giovinazzo 2011; Heslin 1998; Jeekel 1992; Kawai 2006; Lim 2013; Mehta 2013; Paulus 2012; Van Buren 2014). The primary reasons for placing abdominal drains after pancreatic resections are: (ⅰ) drainage of established intra‐abdominal dirty collections (for example bile, pancreatic juice, pus); (ⅱ) prevention of further fluid accumulation; (ⅲ) identification and monitoring of any fistula or bleeding (Adham 2013; Bassi 2010; Conlon 2001; Correa‐Gallego 2013; Fisher 2011; Giovinazzo 2011; Heslin 1998; Jeekel 1992; Kawai 2006; Lim 2013; Mehta 2013; Paulus 2012; Van Buren 2014). Theoretically, abdominal drainage has the potential to prevent or control postoperative complications (e.g., intra‐abdominal abscess, pancreatic or biliary fistula, bleeding; Adham 2013; Bassi 2010; Conlon 2001; Correa‐Gallego 2013; Fisher 2011; Giovinazzo 2011; Heslin 1998; Jeekel 1992; Kawai 2006; Lim 2013; Mehta 2013; Paulus 2012; Van Buren 2014). The use of surgical drains has been considered mandatory after pancreatic surgery for decades (Allen 2011).
However, some surgeons have argued that abdominal drainage may fail to reduce postoperative complications because a drain may become sealed off and ineffective within a few days after pancreatic surgery (Heslin 1998; Paulus 2012). The drain itself appears to act as a foreign body and may interfere with wound healing (Correa‐Gallego 2013; Fisher 2011; Paulus 2012). The drainage tube creates a pathway for contamination and may increase the risk of postoperative infectious complications (Inoue 2011; Jeekel 1992). In addition, the use of a drain may be associated with an increased length of hospital stay (Fisher 2011; Mehta 2013; Paulus 2012). Moreover, abdominal drainage may be associated with some rare adverse events such as bowel perforation, hernia, and bleeding (Cameron‐Strange 1985; Henkus 1999; Makama 2010; Nomura 1998; Reed 1992; Sahu 2008; Srivastava 2007; Van Hee 1983). Recent studies have suggested that routine placement of prophylactic abdominal drains may be unnecessary and may be associated with an increased complication rate (Adham 2013; Correa‐Gallego 2013; Fisher 2011; Giovinazzo 2011; Heslin 1998; Jeekel 1992; Lim 2013; Mehta 2013; Paulus 2012).
Why it is important to do this review
Routine use of prophylactic (protective or preventive) abdominal drainage in patients undergoing pancreatic surgery is controversial. Up to now, there has been no Cochrane review assessing the role of prophylactic abdominal drainage for pancreatic surgery.
Objectives
Primary objective
To assess the benefits and harms of routine abdominal drainage after pancreatic surgery.
Secondary objectives
To determine the comparative effects of different types of surgical drains after pancreatic surgery.
To determine the optimal time for drain removal after pancreatic surgery.
Methods
Criteria for considering studies for this review
Types of studies
We included all randomized controlled trials (RCTs), regardless of sample size, language, or publication status, that compared (ⅰ) drain use and no drain use, (ⅱ) different types of drains, or (ⅲ) different schedules for drain removal in patients undergoing pancreatic surgery. Quasi‐randomized trials, in which the allocation was performed on the basis of a pseudo‐random sequence, e.g., odd/even hospital number or date of birth, alternation, and non‐randomized studies were excluded because of the potential for bias (Reeves 2011).
Types of participants
We included all patients, regardless of age, sex, or race, who underwent elective pancreatic resections (open or laparoscopic) for any pancreatic or extra‐pancreatic disease.
Types of interventions
-
Drain use versus no drain use.
-
One type of drain versus another.
-
Early versus late drain removal (no more than four days versus more than four days).
Types of outcome measures
Primary outcomes
-
Mortality:
-
30‐day mortality;
-
90‐day mortality.
-
-
Infectious complications:
-
intra‐abdominal infection;
-
wound infection.
-
-
Drain‐related complications.
Secondary outcomes
-
Morbidity as defined by study authors. We classified morbidity by the Clavien‐Dindo classification of surgical complications (Clavien 2009).
-
Length of hospital stay.
-
Hospital costs.
-
Additional procedures for postoperative complications:
-
open procedures;
-
radiologic interventions (radiological drainage requiring insertion of drain or percutaneous aspiration)
-
-
Pain, quality of life.
Main outcomes for ‘Summary of findings’ table
-
Mortality.
-
Intra‐abdominal infection.
-
Drain‐related complications.
-
Morbidity.
-
Length of hospital stay.
Search methods for identification of studies
We designed the search strategy with the help of Racquel Simpson (Trials Search Co‐ordinator, Cochrane Upper GI and Pancreatic Diseases Group). We conducted the searches regardless of language, year, or publication status.
Electronic searches
We searched the following to identify randomized controlled trials (Royle 2003):
-
The Cochrane Library (Issue 3, 2015) (Appendix 1),
-
MEDLINE (Ovid) (1946 to 9 April 2015) (Appendix 2),
-
EMBASE (Ovid) (1980 to 9 April 2015) (Appendix 3),
-
Science Citation Index Expanded (Web of Science) (1900 to 9 April 2015) (Appendix 4), and
-
Chinese Biomedical Literature Database (CBM) (1978 to 9 April 2015) (Appendix 5)
Searching other resources
We searched the following databases to identify ongoing trials (accessed 9 April 2015):
-
World Health Organization International Clinical Trials Registry Platform search portal (http://apps.who.int/trialsearch/).
-
ClinicalTrials.gov (http://www.clinicaltrials.gov/).
-
Current Controlled Trials (http://www.controlled‐trials.com/).
-
European (EU) Clinical Trials Register (https://www.clinicaltrialsregister.eu/).
-
Chinese Clinical Trial Register (http://www.chictr.org/).
We also searched the reference lists of identified studies and meeting abstracts via the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES;http://www.sages.org/; accessed 9 April 2015) and Conference Proceedings Citation Index to explore further relevant clinical trials.
Data collection and analysis
We conducted this systematic review according to the Cochrane Handbook for Systematic Reviews of Intervention (Higgins 2011a) and the Cochrane Upper GI and Pancreatic Diseases (UGPD) Group Module (Forman 2011).
Selection of studies
After completing the searches, we merged the search results, using the reference management software package Endnote X7 and removed duplicate records of the same report. Two independent review authors (Wu S, Lu J) scanned the title and abstract of every record identified by the search for inclusion. We retrieved the full text for further assessment if the inclusion criteria were unclear from the abstract (Cheng 2013b). We excluded papers that did not meet the inclusion criteria and listed the reasons for their exclusion.
Data extraction and management
Two authors (Wu S, Lu J) independently extracted, checked, and entered the data into a Microsoft Word electronic data collection form. We resolved any discrepancy between the two authors by discussion.
Assessment of risk of bias in included studies
Two review authors (Xiong X, Ye H) independently assessed the risk of bias in the included trials. We adopted the quality criteria recommended by the Cochrane Handbook for Systematic Reviews of Intervention, version 5.1.0 (Higgins 2011b). The risk of bias of the trials was assessed based on the following domains: random sequence generation; allocation concealment; blinding of outcome assessment; incomplete outcome data; selective reporting; and other sources of bias (Gurusamy 2009; Higgins 2011b). Following the evaluation of these domains, an included trial was judged to be at a low risk of bias if the risk of bias was evaluated as 'low risk' in all of the domains. If the risk of bias of any domain was judged as 'unclear risk' or 'high risk', the trial was listed as 'high risk of bias'. We resolved any difference in opinion by discussion. In cases of unsettled disagreements, a third review author (Cheng N) adjudicated. We presented the results of the risk of bias in two figures (a 'Risk of bias graph' and a 'Risk of bias summary'), which were generated using Review Manager 5 (RevMan 2014).
Measures of treatment effect
We performed the meta‐analyses using Review Manager 5 (RevMan 2014). For dichotomous outcomes, we calculated the risk ratio (RR) with 95% confidence interval (CI; Deeks 2011). For continuous outcomes, we calculated the mean difference (MD) with 95% CI (Deeks 2011).
Unit of analysis issues
The unit of analysis was the individual patient.
Dealing with missing data
We contacted the original investigators to request further information in cases of missing data. However, there was no reply. Thus, we used only the available data in the analyses.
Assessment of heterogeneity
We described the heterogeneity in the data using the Chi² test (Deeks 2011). We considered a P value less than 0.10 to be statistically significant heterogeneity (Deeks 2011). We also used the I² statistic to measure the quantity of heterogeneity. In case of heterogeneity, we performed the meta‐analysis and interpreted the result cautiously.
Assessment of reporting biases
We did not perform funnel plots to assess reporting biases because the number of trials included was less than ten (Sterne 2011).
Data synthesis
We performed the meta‐analyses using Review Manager 5 software provided by The Cochrane Collaboration (RevMan 2014). Two review authors (Lin Y, Zhou R) independently checked and entered all of the data into Review Manager 5. We resolved any discrepancy between the two authors by discussion. For all analyses, we employed the random‐effects model.
Subgroup analysis and investigation of heterogeneity
We had intended to perform the following subgroup analyses, but were unable to because of limited data:
-
RCTs with low risk of bias versus RCTs with high risk of bias.
-
Different etiologies (pancreatic cancer, chronic pancreatitis, and others).
-
The type of operation (proximal, distal, and central pancreatectomy).
Sensitivity analysis
We performed sensitivity analyses to determine whether the conclusions were robust according to the decisions made during the review process by:
-
Changing between a fixed‐effect model and a random‐effects model.
-
Changing between risk ratios (RR), risk differences (RD), and odds ratios (OR) for dichotomous outcomes.
-
Changing between mean differences (MD) and standardized mean differences (SMD) for continuous outcomes.
-
Changing between worst‐case scenario analysis and best‐case scenario analysis for missing data.
If the results did not change, they were considered to have low sensitivity. If the results changed, they were considered to have high sensitivity.
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
Results of the search
We identified a total of 1987 records through the electronic searches of The Cochrane Library (N = 96), MEDLINE (Ovid) (N = 564), EMBASE (Ovid) (N = 548), Science Citation Index Expanded (Web of Science) (N = 555), and Chinese Biomedical Literature Database (CBM) (N = 224). We identified two records through scanning reference lists of the identified randomized controlled trials (Heslin 1998; Jeekel 1992). We excluded 459 duplicates and 1514 clearly irrelevant records through reading titles and abstracts. The remaining 14 records were retrieved for further assessment. We excluded eleven studies for the reasons listed under the table 'Characteristics of excluded studies'. In total, three randomized controlled trials fulfilled the inclusion criteria. The study flow diagram is shown in Figure 1.
Study flow diagram.
Included studies
Drain use versus no drain use
Two trials randomized 316 participants who underwent elective pancreatic resections (276 pancreaticoduodenectomy and 40 distal pancreatectomy) to those who had drainage tubes inserted postoperatively (N = 156) and those who did not (N = 160). Both studies were conducted in the United States of America. The average age was 64.4 years. One or two drainage tubes were placed near both the biliary and pancreatic anastomoses. Both trials measured mortality, morbidity, wound infection, intra‐abdominal infection, various postoperative complications, reoperation, additional radiologic intervention, and length of hospital stay (Conlon 2001; Van Buren 2014).
One type of drain versus another
We did not identify any randomized controlled trials that compared different types of drains.
Early versus late drain removal
One study randomized 114 participants with low risk of postoperative pancreatic fistula undergoing elective pancreatic resections (75 pancreaticoduodenectomy and 39 distal pancreatectomy) to the early drain removal group or the late drain removal group. This trial was conducted in Italy. Two drainage tubes were placed near both the biliary and pancreatic anastomoses. One drainage tube was placed near pancreatic stump after distal pancreatectomy. The average age was 56.6 years. The outcomes reported were pancreatic fistula, abdominal complications, pulmonary complications, reoperation, length of hospital stay, hospital readmission, postoperative mortality, morbidity, and hospital costs.
Excluded studies
We excluded 11 studies. Details are listed in the table 'Characteristics of excluded studies'. One trial was excluded because it focused on pancreatic duct drainage (Lee 2009); the rest were not randomized controlled trials.
Risk of bias in included studies
The risk of bias of the included studies is shown in Figure 2 and Figure 3. All three trials were considered to be of high risk of bias (Bassi 2010; Conlon 2001; Van Buren 2014).
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Random sequence generation was at low risk of bias in two trials (Bassi 2010; Van Buren 2014). Allocation concealment was at unclear risk of bias in all three trials.
Blinding
Blinding of outcome assessment was of high risk of bias in two trials (Bassi 2010; Van Buren 2014).
Incomplete outcome data
Incomplete outcome data was of high risk of bias in one trial (Van Buren 2014).
Selective reporting
The trial protocol was available for two trials (Bassi 2010; Van Buren 2014). All of the studies' pre‐specified outcomes were reported. Thus, both trials were considered to be free of selective reporting. Another trial reported all of the primary outcomes of this review (Conlon 2001). There was some selective outcome reporting in the secondary outcomes, but the review authors considered the trial to be free of selective reporting for the primary outcomes.
Other potential sources of bias
Baseline imbalance was at low risk of bias in all three trials.
Effects of interventions
See: Summary of findings for the main comparison Drain use versus no drain use for pancreatic surgery; Summary of findings 2 Early versus late drain removal for pancreatic surgery
See: summary of findings Table for the main comparison; summary of findings Table 2.
Drain use versus no drain use
Two trials (316 participants) compared drain use with no drain use (Conlon 2001; Van Buren 2014). One hundred and fifty‐six were randomized to the drainage group and 160 patients to the no drainage group. See: summary of findings Table for the main comparison.
Mortality (30 days) (Analysis 1.1)
Both trials (316 participants) reported this outcome. The overall 30‐day mortality was 2.5% (1.3% drain use versus 3.8% no drain use). There was no significant difference in the 30‐day mortality between the groups (RR 0.44; 95% CI 0.05 to 3.94; P = 0.46; heterogeneity: I² = 40%; P = 0.20).
Mortality (90 days) (Analysis 1.2)
Only one trial (137 participants) reported this outcome. The overall 90‐day mortality was 7.3% (2.9% drain use versus 11.6% no drain use). There was no significant difference in the 90‐day mortality between the groups (RR 0.25; 95% CI 0.06 to 1.15; P = 0.08).
Intra‐abdominal infection (Analysis 1.3)
Both trials (316 participants) reported this outcome. The overall intra‐abdominal infection rate was 11.4% (8.3% drain use versus 14.4% no drain use). There was no significant difference in the intra‐abdominal infection rate between the groups (RR 0.61; 95% CI 0.25 to 1.46; P = 0.27; heterogeneity: I² = 41%; P = 0.19).
Wound infection (Analysis 1.4)
Both trials (316 patients) reported this outcome. The overall wound infection rate is 11.4% (10.9% drain use versus 11.9% no drain use). There was no significant difference in the wound infection rate between the groups (RR 0.91; 95% CI 0.45 to 1.86; P=0.80) (heterogeneity: I²=22%; P=0.26).
Drain‐related complications (Analysis 1.5)
Only one trial (179 patients) reported this outcome. There was one drain‐related complication (broken drain) in the drainage group (Conlon 2001).
Morbidity (Analysis 1.6)
Both trials (316 participants) reported this outcome. The morbidity was 66.1% (67.3% drain use versus 65.0% no drain use). There was no significant difference in the morbidity between the groups (RR 1.02; 95% CI 0.88 to 1.19; P = 0.79; heterogeneity: I² = 0%; P = 0.46).
Length of hospital stay (Analysis 1.7)
Both trials (316 participants) reported this outcome. The length of hospital stay was shorter in the drain group (8.1 days) than in the no drain group (8.6 days; MD ‐0.97 days; 95% CI ‐1.41 to ‐0.53; P < 0.0001; heterogeneity: I² = 0%; P = 0.45). The data for length of hospital stay were skewed. The lack of normality for this outcome measure might introduce bias in this review.
Hospital costs
Neither of the trials reported this outcome.
Additional open procedures for postoperative complications (Analysis 1.8)
Both trials (316 participants) reported this outcome. A total of 20 participants need additional open procedures for postoperative complications (10 drain use versus 10 no drain use). There was no significant difference for this outcome between the groups (RR 0.90; 95% CI 0.15 to 5.32; P = 0.91; heterogeneity: I² = 70%; P = 0.07).
Additional radiological interventions for postoperative complications (Analysis 1.9)
Both trials (316 participants) reported this outcome. A total of 40 participants need additional radiological interventions for postoperative complications (17 drain use versus 23 no drain use). There was no significant difference for this outcome between the groups (RR 0.78; 95% CI 0.19 to 3.25; P = 0.74; heterogeneity: I² = 80%; P = 0.02).
Pain/Quality of life
Neither of the trials reported this outcome.
Early versus late drain removal
Only one trial (114 participants with low risk of postoperative pancreatic fistula) was identified that compared early versus late drain removal (Bassi 2010). Fifty‐seven patients were randomized to the early drain removal group and 57 patients to the late drain removal group. See: summary of findings Table 2.
Mortality (30 days) (Analysis 2.1)
There was no mortality in either group.
Intra‐abdominal infection
The trial did not report this outcome
Wound infection
The trial did not report this outcome
Drain‐related complications
The trial did not report this outcome.
Morbidity (Analysis 2.2)
The morbidity was 50% (38.5% early drain removal versus 61.4% late drain removal). The morbidity was lower in the early drain removal group than in the late drain removal group (RR 0.63; 95% CI 0.43 to 0.93; P = 0.02).
Length of hospital stay (Analysis 2.3)
The length of hospital stay was shorter in the early drain removal group (8.7 days) than in the late drain removal group (11.3 days) (MD ‐2.10 days; 95% CI ‐4.17 to ‐0.03; P=0.05).
Hospital costs (Analysis 2.4)
The total hospital costs were lower in the early drain removal group (10,071 Euros) than in the late drain removal group (12,140 Euros) (MD ‐2069.00 Euros; 95% CI ‐3872.26 to ‐265.74; P = 0.02).
Additional open procedures for postoperative complications (Analysis 2.5)
A total of one participant need additional open procedure for postoperative complications (0 early drain removal versus 1 late drain removal). There was no significant difference for this outcome between the groups (RR 0.33; 95% CI 0.01 to 8.01; P = 0.50).
Additional radiological interventions for postoperative complications
The trial did not report this outcome.
Pain/Quality of life
The trial did not report these outcomes.
Subgroup analysis
We did not perform any planned subgroup analysis because only three trials were included in this review.
Sensitivity analysis
We performed the following planned sensitivity analyses:
-
Changing between a fixed‐effect model and a random‐effects model.
-
Changing statistics among risk ratios (RR), risk differences (RD), and odds ratios (OR) for dichotomous outcomes.
-
Changing statistics between mean differences (MD) and standardized mean differences (SMD) for continuous outcomes.
-
Changing between worst‐case scenario analysis and best‐case scenario analysis for missing data.
We observed no change in the results by changing between a fixed‐effect and a random‐effects model, calculating the risk differences (RD) and odds ratios (OR) for dichotomous outcomes, or calculating the SMD for continuous outcomes except for the outcome 'length of hospital stay' (Analysis 3.1). There were three post‐randomization drop‐outs in one trial. We observed no change in the results by changing between worst‐case scenario analysis and best‐case scenario analysis for missing data except for the outcome 'mortality (90 days)'.
Discussion
Summary of main results
For the comparison of drain use with no drain use, we found that there were no clear differences in any of the primary outcomes measured. Drains may lead to drain‐related complications. Data comparing early versus late drain removal were available from only one trial. We found that they favoured early drain removal for patients with low risk of postoperative pancreatic fistula.
The routine use of drains has been considered surgical dogma after a pancreatic resection. Jeekel et al first challenged the dogma three decades ago (Jeekel 1992). They reported a cohort in which 22 patients underwent pancreaticoduodenectomy without undue complications after abandoning abdominal drainage. They concluded that abdominal drainage after pancreaticoduodenectomy may be omitted. Since then, several non‐randomized studies (Adham 2013; Correa‐Gallego 2013; Fisher 2011; Giovinazzo 2011; Heslin 1998; Kawai 2006; Lim 2013; Mehta 2013; Paulus 2012; Witzigmann 2007) and a randomized controlled trial (Conlon 2001) have tested drain use versus no drain use after pancreatic resections. All of the trials found a similar or higher complication rate in the drainage group over the no drainage group. In addition, the routine use of surgical drains was associated with an increased length of hospital stay (Fisher 2011; Mehta 2013; Paulus 2012). Some authors suggested that routine prophylactic drainage after pancreatic resections could be safely abandoned (Adham 2013; Correa‐Gallego 2013; Mehta 2013). Twenty‐two years after the first reports of a 'no drain' policy and 13 years after the first randomized trial (Conlon 2001), Van Buren et al conducted another randomized controlled trial on this topic (Van Buren 2014). They found that pancreaticoduodenectomy without drain use was associated with an increased rate and severity of complications. They questioned the safety of not routinely placing drains in all patients after pancreaticoduodenectomy.
The 30‐day mortality in this review was less than 3% (approximately 2.5%). The 90‐day mortality in this review was approximately 7.5%. There was no significant difference in either 30‐day or 90‐day mortality between the drainage group and the no drainage group. However, this review involved only 316 participants and therefore, was underpowered to detect a significant difference in this outcome.
Surgical morbidity was used to assess the efficacy of prophylactic abdominal drainage to reduce postoperative complications. The routine use of abdominal drainage after pancreatic surgery did not significantly reduce the incidence of postoperative complications (67.3% drain use versus 65.0% no drain use) in this review. There are several possible reasons that may help to explain this. First, surgical drains may reduce some postoperative complications, but they may also be associated with drain‐related postoperative complications (e.g., broken drain). Second, this review only included two trials with 316 participants that compared drain use with no drain use. These studies may not have had the statistical power to detect the efficacy of abdominal drainage to reduce postoperative complications. Interestingly, surgical morbidity was lower in the early drain removal group (38.5%) than in the late drain removal group (61.4%) for patients with low risk of postoperative pancreatic fistula. Thus, surgical drains may not benefit patients with a low risk of postoperative pancreatic fistula. Surgical drains are more likely to increase the incidence of some postoperative complications if the duration of abdominal drainage is prolonged. Many factors have been considered to influence the development of postoperative pancreatic fistula (e.g., age, obesity, cardiovascular diseases, diabetes mellitus, pancreatic texture, pancreatic duct size; Ramacciato 2011). It seems that older (e.g., over 60 years of age), overweight people with cardiovascular diseases, diabetes mellitus, soft pancreatic texture, and a small pancreatic duct diameter (e.g., less than 3 mm) are more likely to suffer postoperative pancreatic fistula (Ramacciato 2011; Riall 2008). A recent RCT demonstrated that the use of a somatostatin analogue reduced the risk of postoperative pancreatic fistula (Allen 2014). The use of drains may be more relevant in patients with a higher risk of postoperative pancreatic fistula, which requires further investigation.
Length of hospital stay and hospital costs are important outcomes from the patient and healthcare funder perspectives. In the case of drain insertion, early drain removal was associated with a shorter length of hospital stay (2.10 days) and lower hospital costs (2069 Euros) than a later drain removal. There were no randomized trials that compared an open drain with a closed one. Thus, we were unable to draw any conclusion about the optimal type of drain in cases where the surgeon considered abdominal drainage is necessary.
Overall completeness and applicability of evidence
All of the trials included patients undergoing elective pancreaticoduodenectomy (N = 351, 81.6%) and distal pancreatectomy (N = 79, 18.4%) for various pancreatic and extra‐pancreatic diseases, including pancreatic cancers, ampullary cancers, chronic pancreatitis, biliary and duodenal malignancy, etc. The majority (72.7%) of the patients had pancreatic cancers (61.6%) or ampullary cancers (11.1%). Because only one trial included 11 patients (2.6%) undergoing laparoscopic pancreatic resections, the results of this review are not applicable to patients undergoing laparoscopic pancreatic resections. The role of abdominal drainage after laparoscopic pancreatic resections therefore requires further assessment. Thus, the results of this review are only applicable to patients undergoing elective open pancreaticoduodenectomy and distal pancreatectomy for various pancreatic and extra‐pancreatic diseases, especially for pancreatic cancers or ampullary cancers.
Quality of the evidence
None of the trials were at low risk of bias. The trials included under each comparison were too few to assess inconsistency and publication bias. There was no indirectness of evidence because the trials did not perform the indirect comparison of one type of drain versus another. The confidence intervals of the majority of outcomes were wide, indicating that the estimates of effect obtained were imprecise. Overall, the quality of the evidence was considered to be very low (summary of findings Table for the main comparison; summary of findings Table 2).
Potential biases in the review process
There were only three trials with 430 patients included in the review; thus, there are a lack of data on this topic to date. This review is subject to both alpha error (false positivity) and beta error (false negativity). Additionally, we did not perform funnel plots to assess the publication bias due to the small number of included trials.
Agreements and disagreements with other studies or reviews
There is increasing evidence in Cochrane reviews that routine abdominal drainage after various abdominal operations is not mandatory (Cheng 2015; de Jesus 2004; Gurusamy 2007a; Gurusamy 2007b; Gurusamy 2007c; Wang 2011). The routine use of surgical drains has also been questioned in other areas, including thyroid, gynaecological, and orthopedic surgeries (Charoenkwan 2010; Gates 2005; Parker 2007; Samraj 2007).
One systematic review that compared drain use with no drain use in patients undergoing pancreatic resections included three trials that we had considered for this review (Conlon 2001; Fisher 2011; Heslin 1998). Two of these trials were non‐randomized, so were not included in this review They concluded that the routine use of abdominal drains after pancreatic resection may result in a higher risk for major complications (Van der Wilt 2013). This review does not make any specific recommendation because the number of the participants included in this review does not have the statistical power to detect the benefit of abdominal drainage. A sample size of 870 (435 in each group) would be required to detect an absolute reduction in the intra‐abdominal infection rate of 5% (from 10% to 5%) at 80% power and an alpha‐error set at 0.05 (Conlon 2001).
Another systematic review compared early drain removal with late drain removal in patients undergoing pancreatic resection and included two trials that we had considered for this review (Bassi 2010; Kawai 2006). One of the trials was not randomized, so we did not include it. They concluded that early drain removal seemed to be superior to late drain removal (Diener 2011), while our results concluded that early drain removal seemed to be superior to late drain removal for patients with low risk of postoperative pancreatic fistula.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Drain use versus no drain use, Outcome 1 Mortality (30 days).
Comparison 1 Drain use versus no drain use, Outcome 2 Mortality (90 days).
Comparison 1 Drain use versus no drain use, Outcome 3 Intra‐abdominal infection.
Comparison 1 Drain use versus no drain use, Outcome 4 Wound infection.
Comparison 1 Drain use versus no drain use, Outcome 5 Drain‐related complications.
Comparison 1 Drain use versus no drain use, Outcome 6 Morbidity.
Comparison 1 Drain use versus no drain use, Outcome 7 Length of hospital stay.
Comparison 1 Drain use versus no drain use, Outcome 8 Additional open procedures for postoperative complications.
Comparison 1 Drain use versus no drain use, Outcome 9 Additional radiological interventions for postoperative complications.
Comparison 2 Early versus late drain removal, Outcome 1 Mortality (30 days).
Comparison 2 Early versus late drain removal, Outcome 2 Morbidity.
Comparison 2 Early versus late drain removal, Outcome 3 Length of hospital stay.
Comparison 2 Early versus late drain removal, Outcome 4 Hospital costs (Euros).
Comparison 2 Early versus late drain removal, Outcome 5 Additional open procedures for postoperative complications.
Comparison 3 Drain use versus no drain use (sensitivity analysis by calculating the SMD for length of hospital stay), Outcome 1 Length of hospital stay.
Comparison 4 Drain use versus no drain use (sensitivity analysis by changing between worst‐case scenario analysis and best‐case scenario analysis for missing data), Outcome 1 Mortality (90 days).
Comparison 4 Drain use versus no drain use (sensitivity analysis by changing between worst‐case scenario analysis and best‐case scenario analysis for missing data), Outcome 2 Mortality (90 days).
| Drain use versus no drain use for pancreatic surgery | ||||||
| Patient or population: Patients undergoing elective open pancreatic resections | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| No drain use | Drain use | |||||
| Mortality (30 days) | Study population | RR 0.44 | 316 | ⊕⊝⊝⊝ | ||
| 38 per 1000 | 16 per 1000 | |||||
| Moderate | ||||||
| 40 per 1000 | 18 per 1000 | |||||
| Mortality (90 days) | Study population | RR 0.24 | 134 | ⊕⊝⊝⊝ | ||
| 121 per 1000 | 29 per 1000 | |||||
| Moderate | ||||||
| 121 per 1000 | 29 per 1000 | |||||
| Intra‐abdominal infection | Study population | RR 0.61 | 316 | ⊕⊝⊝⊝ | ||
| 144 per 1000 | 88 per 1000 | |||||
| Moderate | ||||||
| 156 per 1000 | 95 per 1000 | |||||
| Drain‐related complications | Study population | RR 3.1 | 179 | ⊕⊝⊝⊝ | There was one drain‐related complication in the drainage group. The drainage tube was broken. | |
| 0 per 1000 | 12 per 1000 | |||||
| Moderate | ||||||
| 0 per 1000 | 0 per 1000 | |||||
| Morbidity | Study population | RR 1.02 | 316 | ⊕⊝⊝⊝ | ||
| 650 per 1000 | 663 per 1000 | |||||
| Moderate | ||||||
| 663 per 1000 | 676 per 1000 | |||||
| Length of hospital stay | The mean length of hospital stay in the no drain group was 8.12 days | The mean length of hospital stay in the drain groups was | MD ‐0.97 (‐1.41 to ‐0.53) | 316 | ⊕⊝⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level for serious risk of bias. 5 Downgraded one level for serious imprecision (Total population size was less than 400). | ||||||
| Early versus late drain removal for pancreatic surgery | ||||||
| Patient or population: Patients undergoing elective open pancreatic resections | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Late drain removal | Early drain removal | |||||
| Mortality (30 days) | See comment | See comment | Not estimable | 114 | ⊕⊕⊝⊝ | There was no mortality in either group. |
| Morbidity | Study population | RR 0.63 | 114 | ⊕⊕⊝⊝ | ||
| 614 per 1000 | 387 per 1000 | |||||
| Moderate | ||||||
| 614 per 1000 | 387 per 1000 | |||||
| Length of hospital stay (days) | The mean length of hospital stay in the late removal group was 10.8 days | The mean length of hospital stay in the early removal group was | MD ‐2.10 (‐4.17 to ‐0.03) | 114 | ⊕⊕⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level due to serious risk of bias. | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Mortality (30 days) Show forest plot | 2 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 0.44 [0.05, 3.94] |
| 2 Mortality (90 days) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 3 Intra‐abdominal infection Show forest plot | 2 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 0.61 [0.25, 1.46] |
| 4 Wound infection Show forest plot | 2 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.45, 1.86] |
| 5 Drain‐related complications Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 6 Morbidity Show forest plot | 2 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 1.02 [0.88, 1.19] |
| 7 Length of hospital stay Show forest plot | 2 | 316 | Mean Difference (IV, Random, 95% CI) | ‐0.97 [‐1.41, ‐0.53] |
| 8 Additional open procedures for postoperative complications Show forest plot | 2 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 0.90 [0.15, 5.32] |
| 9 Additional radiological interventions for postoperative complications Show forest plot | 2 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 0.78 [0.19, 3.25] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Mortality (30 days) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 2 Morbidity Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 3 Length of hospital stay Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 4 Hospital costs (Euros) Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 5 Additional open procedures for postoperative complications Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Length of hospital stay Show forest plot | 2 | 316 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.36 [‐1.08, 0.36] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Mortality (90 days) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 2 Mortality (90 days) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
