Common bile duct stones are the second most frequent complication of cholecystolithiasis, occurring in approximately 5 % of asymptomatic patients with a normal diameter bile duct on trans-abdominal ultrasound scan at the time of cholecystectomy, and in 10–20 % of patients with symptomatic gallstones [1, 2]. Treatment is advisable to prevent further complications, such as obstructive jaundice, acute cholangitis, and pancreatitis [3]. The optimal treatment for common bile duct stones is still unclear, and the options available include open common bile duct exploration (CBDE), laparoscopic common bile duct exploration (LCBDE), and pre-, intra- or postoperative endoscopic retrograde cholangiopancreatography with sphincterotomy (ERCP and ES) combined with laparoscopic cholecystectomy [2].

In the era of open cholecystectomy, CBDE was the gold-standard procedure for CBD stones, but nowadays, with advances in laparoscopic instrumentation and acquisition of advanced laparoscopic skills, LCBDE for choledocholithiasis is increasing in popularity among surgeons worldwide [4, 5]. There is evidence in the literature to suggest that LCBDE for choledocholithiasis is of equal efficacy, is associated with equal morbidity rate, and is more cost-effective than ERCP followed by laparoscopic cholecystectomy [6–8].

LCBDE may be performed trans-cystic or by direct choledochotomy, and this is determined by stone size, load and distribution, and also the diameter of the cystic duct [9, 10]. When there is an indication for direct bile duct dochotomy to clear the stone burden, this is subsequently managed by primary duct closure (PDC) or closure with T-tube drainage (TTD). TTD was common practice in open CBD exploration and has been common practice after LCBDE to achieve postoperative decompression of the common bile duct and visualization of the biliary system through cholangiography to check for residual stones [11, 12]. However, this practice is associated with significant T-tube-related complications that include drain site pain, biliary leak, CBD obstruction due to accidental tube dislodgement, persistent biliary fistula, and biliary peritonitis due to tube dislodgement or after T-tube removal. These complications are reported to occur in approximately 15 % of patients with TTD [13, 14]. Furthermore, T-tube insertion after laparoscopic or open CBDE is associated with prolonged hospital stay, longer operating time, and higher hospital expenses [6, 15–19].

Consequently, some surgeons have recommended primary closure of the common bile duct immediately after dochotomy to reduce the risk of T-tube-related complications, and also to facilitate early discharge, early return to normal activity, and less hospital expenses [15, 20, 21].

Various internal and external biliary drainage methods have been analyzed in the literature in order to decompress the biliary tree after LCBDE and primary duct closure + biliary drain (PDC + BD), with ante-grade biliary stents, modified biliary stents, modified intra-cystic biliary catheters, and J-tubes which are all described [22–26]. These authors showed that internal biliary stenting following LCBDE is an effective and safe technique that prevents T-tube-related morbidity and results in a shorter postoperative hospital stay and an earlier return to work, when compared to TTD [27, 28].

To date four meta-analyses have been performed to compare the results of PDC with those of TTD [21, 29–31]. The most complete pooled analysis, performed by Yin et al. [31], enrolled twelve studies (three randomized controlled trials and nine retrospective cohort studies) comparing PDC, with or without BD insertion, and TTD.

This review has included two new randomized controlled trials comparing PDC versus TTD, and PDC + BD insertion versus TTD [18, 28]. Moreover, this review has included five retrospective cohort studies in the pooled analysis that have not been included previously [4, 17, 27, 32, 33]. Therefore, this meta-analysis reports on the largest number of patients from all randomized controlled trials and retrospective cohort studies in the literature to assess and validate the safety, feasibility, and potential benefits or limitations of PDC when compared to TTD after LCBDE.

Materials and methods

Search methods for identification of studies

A systematic literature search was performed using PubMed, EMBASE, MEDLINE, Google Scholar, and the Cochrane Central Register of Controlled Trials databases for studies comparing PDC and TTD. We combined database-specific search terms for primary closure (primary duct closure or primary closure or primary suture or PDC), T-tube (T-tube or T-tube drainage), and LCBDE (laparoscopic common bile duct exploration, laparoscopic choledochotomy). The search was then extended to related articles suggested by the databases and supplemented with manual searches for reference lists of all relevant articles. When the results of a single study were reported in more than one publication, only the most recent and complete data were included in the meta-analysis. Literature search was completed in September 2014.

Selection of studies

This systematic review and meta-analysis included randomized controlled trials (RCTs) and retrospective cohort studies (RCSs), in which different techniques of PDC and TTD after LCBDE were compared, irrespective of language, blinding, or publication status. To be included in the analysis, studies had to meet the following inclusion criteria:

  1. a.

    Patients did not have any contraindication for laparoscopic surgery.

  2. b.

    Patients did not have acute biliary pancreatitis, ampullary stenosis with multiple intrahepatic stones, severe acute cholangitis, suspected biliary neoplasia, hemorrhagic tendency due to any reason, known cirrhosis of the liver.

  3. c.

    The included studies were required to report at least one of the following outcomes measures of the different techniques used for treatment: postoperative overall morbidity, postoperative biliary-specific complications, re-intervention rate, operating time, postoperative hospital stay, or median hospital expenses.

The exclusion criteria were: articles not reporting data on the outcomes of interest or articles in which the outcomes of interest were impossible to calculate, non-human studies, review articles, editorials, letters and case reports.

Types of outcome measures

Primary outcomes assessed were: overall postoperative complications, postoperative biliary-specific complications (biliary peritonitis, biliary leak, retained stones and postoperative common bile duct obstruction), re-intervention (radiology/endoscopy), re-intervention (surgery) and postoperative hospital stay.

Secondary outcomes assessed were: operating time, median hospital expenses, and other general complications not directly related to the techniques of bile duct closure (wound infection, pneumonia, deep vein thrombosis, internal hemorrhage).

Data extraction and management and assessment of risk of bias in included studies

Two reviewers (M.P and I.S.T) independently considered the eligibility of potential titles and extracted data. Discrepancies were resolved by mutual discussion. Inclusion and exclusion criteria, country and year of publication, study type, number of patients operated on with each technique, and the general characteristics of patients (age, gender, perioperative outcome, postoperative results) were extracted. The risk of bias for the trials enrolled in the meta-analysis were evaluated according to the Cochrane Handbook for Systematic Reviews of Interventions, while the quality of non-randomized studies was assessed using the criteria suggested by the Newcastle–Ottawa quality assessment tool [34, 35]. According to this scale, the maximum score could be nine points, representing the highest methodological quality.

Data synthesis

Systematic review with meta-analysis was performed in accordance with the recommendations from the preferred items for systematic reviews and meta-analyses statement (PRISMA) [36]. The effect sizes were calculated by odds ratio (OR) for dichotomous variables and weighted mean differences (WMD) for continuous outcome measures with 95 % confidence intervals (CIs). The point estimate of the OR value was considered statistically significant at P level of less than 0.05 if the 95 % CI did not cross the value 1. The point estimate of the WMD value was considered statistically significant at P level of less than 0.05 if the 95 % CI did not cross the value 0. Heterogeneity of the results across studies was assessed using the Higgins’ I 2 and Chi-square tests.

A P value of Chi-square test less than 0.10 with an I 2 value of greater than 50 % were considered as indicative of substantial heterogeneity [34]. Fixed-effects model was applied if statistically significant heterogeneity was absent; otherwise, a random-effects model was used for meta-analysis if statistically significant heterogeneity was found, according to the method of DerSimonian and Laird [37]. Statistical analysis was performed using Review Manager software [38].

Results

Description of studies

A total of 315 references were identified through electronic database searches. 290 searches were excluded based on titles and abstract reviews because they did not match the inclusion criteria of the meta-analysis or they reported data from open choledochotomy. The remaining 25 publications underwent full article review. A further eight publications were excluded because they did not focus on the subject. One prospective randomized trial [39] was excluded because it showed the preliminary data of another study [18]. A total of sixteen studies comparing PDC and TTD qualified for inclusion in this review and meta-analysis. Four were randomized controlled trials [15, 18, 19, 28] and twelve were retrospective cohort studies [4, 16, 17, 22–27, 32, 33, 40], with a total of 1770 patients: 1012 in the subgroup analysis PDC versus TTD and 758 patients in the subgroup analysis PDC + BD insertion versus TTD (Fig. 1). The meta-analysis performed by Yin et al. [31] included 956 patients. However, this pooled analysis contained the patients enrolled from the study by Fujimura et al. [41] which included open procedures during his early experience. Therefore, this study was excluded, and the patients analyzed by Martin et al. [4] Morcillo et al. [33], Cai et al. [32], Zhang et al. [17] and Martinez-Baena et al. [27] were included. Two new randomized controlled trials which were not analyzed in previous meta-analyses were: the RCT performed by Zhang HW et al. which compared PDC and TTD and the RCT published by Mangla et al. which reported a comparison between PDC + BD insertion and TTD [17, 28].

Fig. 1
figure 1

The PRISMA flow chart for systematic search and selection of articles for review and meta-analysis

Full size image

Table 1 shows the characteristics of the included studies and the demographic details regarding the enrolled patients. The sixteen articles involved in the quantitative synthesis were published between 1998 and 2014.

Table 1 Characteristics of the included studies and demographic characteristics of the patients (PDC vs. TTD and PDC + BD vs. TTD)
Full size table

Table 2 lists the methods for PDC, PDC + BD, and TTD. Suture techniques of primary closure included running absorbable sutures or interrupted absorbable sutures. Studies also described the T-tube type. Normally, a latex rubber T-tube (14–20 Fr) was used, and the choledochotomy around the tube was closed both through running and interrupted absorbable sutures. T-tube removal times varied depending on the surgeons’ experience from 8 days minimum [40] to 12 weeks maximum [32]. Various techniques have been described for PDC + BD. Ante-grade biliary stent insertions under direct vision of choledochoscopy are described in six studies [4, 23, 25, 27, 28, 33], modified biliary stents were inserted in one study [22], modified trans-cystic biliary catheter in one study [24], J-tube drainage in one study [26], and preoperative percutaneous transhepatic cholangiographic drainage (PTCD) in another study [40]. All authors reported the use of abdominal drain.

Table 2 Methods for primary duct closure and for T-tube drainage (PDC vs. TTD and PDC + BD vs. TTD)
Full size table

Risk of bias in included randomized controlled trials

The risk of bias in the four randomized controlled trials [15, 18, 19, 28] was assessed through the Cochrane Collaboration Risk of Bias Tool. Results are shown in Table 3. Allocation sequence generation was clearly described by authors in two studies [15, 28], while concealment and blinding of the patient, personnel, and observer were clearly reported in the study by El-Geidie [19]. Adequate assessment of each outcome and selective outcome reporting were determined for all trials, but authors did not report intention-to-treat analysis for outcomes. Power analysis calculation for minimum sample size has not been provided by any author and handling of missing data remained unclear. For the 12 RCSs, risk of bias was evaluated by the Newcastle–Ottawa scale. Two studies achieved five stars, and seven studies achieved four stars. Outcomes may have been influenced by allocation bias in all RCSs for patients who underwent PDC or TTD. Furthermore, the follow-up length was unclear in most of the RCSs.

Table 3 Risk of bias in the published randomized controlled trials (by the Cochrane Risk of Bias Tool) and in the retrospective cohort studies (by the Newcastle–Ottawa quality assessment tool)
Full size table

Effects of interventions

Primary outcome measures

All outcome measures have been evaluated in order to assess the safety and feasibility of PDC, PDC + BD, and TTD. For all the primary outcomes, the detailed results are reported in Table 4 and Fig. 2.

Table 4 Primary outcomes of interest (PDC vs. TTD and PDC + BD vs. TTD)
Full size table
Fig. 2
figure 2

Meta-analysis of primary outcomes of interest. Primary duct closure (PDC) versus T-tube drainage (TTD) and primary duct closure + biliary drainage (PDC+) versus T-tube drainage (TTD)

Full size image

Postoperative overall morbidity

Complications were reported for 36 patients (7.4 %) in the PDC group and for 55 patients (11.6 %) in the TTD group. The overall morbidity rate was slightly lower in the PDC group than in the TTD group, but this difference is not statistically significant (OR 0.64, 95 % CI 0.41–1.00, P = 0.05; no heterogeneity was found for I 2 = 0 %; P = 0.83).

In the subgroup analysis of PDC + BD versus TTD, complications were reported for 33 patients (13.2 %) in the PDC + BD group and for 55 patients in the TTD group (16.2 %). The overall morbidity rate appeared slightly lower in the PDC + BD group than in the TTD group, but this was not a significant difference (OR 0.77, 95 % CI 0.47–1.25, P = 0.29; no heterogeneity was found for I 2 = 16 %; P = 0.30) (Table 4; Fig. 2).

Postoperative biliary-specific complications

In the overall meta-analysis of RCTs and RCSs comparing all biliary-specific complications after PDC and TTD, there was no significant difference in biliary-specific complications. Complications were reported in 28 patients (5.8 %) in the PDC group and in 40 patients (8.4 %) in the TTD group (OR 0.69, 95 % CI 0.42–1.16, P = 0.16; no heterogeneity was found for I 2 = 0 %; P = 0.86).

When comparing the PDC + BD group and the TTD group, biliary-specific complications were again similar (21 cases, 8.4 % versus 40 cases, 11.8 %. OR 0.69, 95 % CI 0.39–1.24, P = 0.22; no heterogeneity was found for I 2 = 0 %; P = 0.56) (Table 4; Fig. 2). However, when the biliary-specific complications are analyzed individually, differences are apparent between the different techniques for dochotomy closure.

  1. 1.

    Biliary peritonitis

Eight studies comparing PDC with TTD provided data on postoperative biliary peritonitis. No events were reported in the PDC group, while 12 events (2.3 %) were reported in the TTD group. So, primary closure showed a lower rate of postoperative biliary peritonitis, with a statistically significant difference (OR 0.22, 95 % CI 0.06–0.76, P = 0.02; no heterogeneity was found for I 2 = 0 %; P = 0.99). There was no statistically significant difference between the PDC + BD group and the TTD group for this outcome of interest. However, the result tended to favor the PDC + BD group (0.3 vs. 3.8 %, OR 0.35, 95 % CI 0.12–1.06, P = 0.06; no heterogeneity was found for I 2 = 0 %; P = 0.88) (Table 5; Fig. 3).

Table 5 Biliary-specific complications (PDC vs. TTD and PDC + BD vs. TTD)
Full size table
Fig. 3
figure 3

Meta-analysis of biliary-specific complications. Primary duct closure (PDC) versus T-tube drainage (TTD) and primary duct closure + biliary drainage (PDC+) versus T-tube drainage (TTD)

Full size image
  1. 2.

    Biliary leak

No statistically significant difference was found regarding postoperative biliary leak in the meta-analysis of studies comparing PDC and TTD. Nineteen cases (3.9 %) were reported in the PDC group, and 17 cases (3.6 %) were reported in the TTD group (OR 1.13, 95 % CI 0.58 to 2.21, P = 0.71; no heterogeneity was found for I 2 = 0 %; P = 0.94).

When comparing PDC + BD and TTD, no statistically significant difference for this outcome was found between the techniques (17 cases, 6.8 % vs. 19 cases, 5.6 %. OR 1.05, 95 % CI 0.53–2.06, P = 0.89; no heterogeneity was found for I 2 = %; P = 0.43) (Table 5; Fig. 3).

  1. 3.

    Retained stones and postoperative common bile duct obstruction

No significant difference was found in the meta-analysis of studies comparing PDC and TTD for retained stones (1.3 vs. 1.4 %, OR 0.95, 95 % CI 0.32–2.87, P = 0.93; no heterogeneity was found for I 2 = 0 %; P = 0.77) or postoperative common bile duct obstruction (0.4 vs. 0.6 %, OR 0.81, 95 % CI 0.16–4.12, P = 0.80; no heterogeneity was found for I 2 = 0 %; P = 0.77).

The meta-analysis of the studies comparing PDC + BD and TTD showed no significant difference for retained stones (1.7 vs. 3.9 %, OR 0.53, 95 % CI 0.18–1.52, P = 0.24; no heterogeneity was found for I 2 = 0 %; P = 0.99). Two cases of postoperative CBD obstruction were reported in each group (0.9 vs. 0.6 %, OR 2.81, 95 % CI 0.58–13.65, P = 0.20; no heterogeneity was found for I 2 = 0 %; P = 0.35) (Table 5; Fig. 3).

Re-intervention: radiology/endoscopy and re-intervention: surgery

The meta-analysis of the data regarding the radiological and/or endoscopic re-interventions showed a similar rate of re-operations in the PDC group and in the TTD group (2.3 vs. 3.4 %, OR 0.69, 95 % CI 0.33–1.45, P = 0.33; no heterogeneity was found for I 2 = 0 %; P = 0.86). PDC + BD showed a similar rate of radiological and/or endoscopic re-intervention when compared to TTD (5 cases, 2.3 vs. 9 cases, 2.9 %, OR 0.71, 95 % CI 0.24–2.15, P = 0.54; no heterogeneity was found for I 2 = 0 %; P = 0.98). (Table 4; Fig. 2).

The surgical re-intervention rate was similar in the PDC group and in the TTD group (1 vs. 2.1 %, OR 0.63, 95 % CI 0.23–1.72, P = 0.37; no heterogeneity was found for I 2 = 32 %; P = 0.22). The meta-analysis of studies comparing PDC + BD and TTD showed a slightly lower rate of surgical re-interventions in the PDC + BD group, but this difference was not statistically significant (1 vs. 4.2 %, OR 0.39, 95 % CI 0.12–1.24, P = 0.11; no heterogeneity was found for I 2 = 0 %; P = 0.70) (Table 4; Fig. 2).

Postoperative hospital stay (days)

PDC was associated with a shorter postoperative hospital stay (WMD, −3.22; 95 % CI −4.52 to −1.92, P < 0.00001; heterogeneity was found for I 2 = 95 %; P < 0.00001). Within the subgroup analysis of studies comparing PDC + BD and TTD, the biliary drain insertion technique showed a shorter postoperative hospital stay (WMD, −2.68; 95 % CI −3.23 to −2.13, P < 0.00001; no heterogeneity was found for I 2 = 0 %; P = 0.52) (Table 4; Fig. 2).

Secondary outcome measures

Other general complications (wound infection, pneumonia, deep vein thrombosis, internal hemorrhage)

No significant difference was found for general complications when comparing PDC and TTD (1 vs. 2.5 %, OR 0.52, 95 % CI 0.21–1.32, P = 0.17; no heterogeneity was found for I 2 = 0 %; P = 0.98). The meta-analysis of the studies comparing PDC + BD and TTD showed no statistically significant difference (3.7 vs. 3.3 %, OR 1.16, 95 % CI 0.44–3.10, P = 0.76; no heterogeneity was found for I 2 = 0 %; P = 0.71) (Table 6; Fig. 4).

Table 6 Secondary outcomes of interest (PDC vs. TTD and PDC + BD vs. TTD)
Full size table
Fig. 4
figure 4

Meta-analysis of secondary outcomes of interest. Primary duct closure (PDC) versus T-tube drainage (TTD) and primary duct closure + biliary drainage (PDC +) versus T-tube drainage (TTD)

Full size image
Operating time (minutes)

Mean operative time was significantly shorter in PDC group than in TTD group (WMD, −22.27, 95 % CI −33.26 to −11.28, P < 0.00001; heterogeneity was found for I 2 = 95 %; P < 0.00001). On the other hand, the pooled analysis of studies comparing PDC + BD and TTD showed no statistically significant difference for this outcome (WMD, −9.96, 95 % CI −22.00 to 2.08, P = 0.10; heterogeneity was found for I 2 = 81 %; P = 0.0003) (Table 6; Fig. 4).

Median hospital expenses

Only two randomized controlled trials comparing PDC and TTD reported data on median hospital expenses that were noted to be less with PDC. According to the Cochrane Consumers and Communication Review Group indications, the pooled analysis of data was feasible [42]. The difference was statistically significant (SMD, −1.37, 95 % CI −1.96 to −0.77, P < 0.00001; heterogeneity was found for I 2 = 77 %; P = 0.04) (Table 6; Fig. 4).

Discussion

Management of choledocholithiasis has changed radically in recent years following innovation and developments in minimally invasive surgical techniques. Consensus on the optimal therapy for the management of common bile duct stones remains unclear. Preoperative endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic sphincterotomy (ES) followed by laparoscopic cholecystectomy is a popular option for the treatment of this disease. Nevertheless, ERCP and ES are associated with biliary complications in 8–10 % of patients [43, 53]. Long-term complications of ES were reported in a study of 310 patients with a median follow-up period of 74 months: 7.4 % of patients had recurrent ductal stones, 1.6 % had cholangitis, 0.6 % had stenosis of the papilla, and 0.3 % had biliary pancreatitis [54].

LCBDE for common bile duct stones is cost-effective and has a similar rate of associated morbidity when compared to the two-stage method of ERCP and ES followed by laparoscopic cholecystectomy [44–46]. The large multicenter randomized controlled trial published by Cuschieri et al. [6] indicated that in fit patients (ASA I and II), single-stage laparoscopic treatment is the better option, whereas acute cholangitis, severe biliary pancreatitis, ampullary stone impaction or severe comorbid disease represented relative contraindications for LCBDE and they should be approached preoperatively through ERCP and ES.

LCBDE may be performed either through the cystic duct or through a choledochotomy.

Whenever feasible, the trans-cystic duct approach is the preferred technique, because it is less invasive and has proved to be safe and efficient [47].

The indications for trans-cystic CBDE, however, are limited to stones that are smaller than the size of the cystic duct, to a number of stones, to stones located in the lower CBD and not higher up in the common hepatic duct, and when a favorable anatomy of the cystic duct-CBD junction is present. On the other hand, a choledochotomy is better indicated when the CBD diameter is larger than 8–10 mm and when any of these conditions are detected at the intraoperative cholangiogram: stones considerably larger than the lumen of the cystic duct; more than five CBD stones; low and medial cystic duct-CBD junction; common hepatic duct stones [55].

Historically, exploration of the CBD with both open and laparoscopic surgery was accompanied by the insertion of a T-tube drain in order to minimize the risk of postoperative complications, decompress the biliary tree, and provide easy percutaneous access for cholangiogram and extraction of retained stones. Moreover, T-tube drainage was considered to be necessary to allow the edema and swelling at the Ampulla of Vater time to recover after the trauma of the surgery. However, complications of T-tube insertion have been reported in the literature with a rate of about 10–15 % [13, 39]. Some of these complications, such as biliary peritonitis after T-tube removal or biliary leak due to tube dislodgement, are serious and can lead to a need for further interventions. Furthermore, the presence of T-tube in situ contributes to delayed return to normal activity and work and may cause patients persistent pain and discomfort [15]. Recent randomized controlled trials and retrospective cohort studies with long follow-up periods have shown that PDC with or without BD after LCBDE is a safe alternative to the insertion of a T-tube [15–19].

The Cochrane intervention review published by Gurusamy et al. [21] in 2013 on data from three RCTs concluded that TTD resulted in significantly longer operating time and hospital stay as compared to PDC, without any evidence of advantage. However, the number of patients included in this study was too small to make a firm practice recommendation.

The overall complication rates between LCBDE with choledochotomy and ERCP were comparable in the randomized controlled trial published by Bansal et al. [46] with the total amount of complications reported in the LCBDE group classified as Clavien-Dindo I, and the complications reported in the ERCP and ES group distributed among all classes.

In our systematic review and meta-analysis, the rate of overall morbidity was found to be slightly lower in the PDC (7.4 %) and PDC + BD (13.2 %) groups than in the TTD group (11.6 and 16.2 %), but this difference was not statistically significant.

On the other hand, our meta-analysis showed a significantly (P = 0.02) lower rate of biliary peritonitis in the PDC group when compared with the TTD group, with no cases reported after PDC and 12 cases (2.3 %) after TTD. Interestingly, 19/21 cases of biliary peritonitis occurred after T-tube removal, probably due to the insufficient adhesions for T-tube tract formation [18, 19, 22]. Yin et al. in their meta-analysis compared the overall and biliary-specific complications in the studies where T-tubes were removed between 8 and 16 days with the studies where T-tubes were removed after more than 21 days, showing a higher rate of biliary-specific complications occurred when the T-tube was removed earlier [31].

Together with the biliary fistula, CBD stricture is the main complication of LCBDE. PDC of choledochotomy in CBD with diameter <7 mm is related to postoperative stricture and therefore is suggested to be safe only if diameter is >7–9 mm [4, 52].

Decker et al. and Cai et al. reported no biliary strictures in their records of, respectively, 100 and 137 choledochotomies, performed through transverse incision and longitudinal incision, respectively, [32, 50]. In our meta-analysis, postoperative CBD obstruction due to postoperative common bile duct strictures and T-tube twisting was 0.4 % (2 cases) for PDC and 0.6 % (3 cases) for TTD group. Two cases for each group were reported when comparing PDC + BD and TTD.

Although retained bile duct stone after LCBDE with choledochotomy has been reported in up to 4 % of cases in the large study by Khaled et al., this meta-analysis showed high rates of CBD clearance, with no statistically significant difference in the incidence of retained stones when comparing the three techniques for dochotomy closure [48].

In this meta-analysis, we have not found any significant advantage when using biliary stent instead of T-tube positioning, with exception of postoperative hospital stay. The same results were reported in 2007 study by Tailor et al. [49] who did not find any benefit or reduction in rates of biliary leak with use of biliary stents after LCBDE. Other authors recommend selective stent use only when purulent material, sludge, or numerous stones have been extracted from the biliary system.

Length of hospital stay may be influenced by factors that are independent of the patient’s postoperative recovery, such as socioeconomic status and local healthcare systems. Nevertheless, patients who underwent PDC had a significantly shorter postoperative hospital stay which is in concordance with many other trials and comparative studies [16, 22, 23, 28, 32, 40]. This meta-analysis also demonstrated a significant reduction in operating time when comparing PDC and TTD.

The longer operating time for the TTD group of patients may have resulted from the complexity of T-tube insertion and subsequent dochotomy closure techniques used. A prolonged operating time and duration of anesthesia are thought to be related to an increased risk of thromboembolic, respiratory, and cardiac complications, as suggested by Wu et al. [30] in their meta-analysis.

An important limitation of this review is the small number of well-designed randomized controlled trails that have reported on this subject to date. Consequently, strong recommendations on the definitive closure technique of a bile duct dochotomy cannot be overly dogmatic until the associations that are described here in this report are validated by further well-constructed and appropriately powered RCTs. Furthermore, missing information regarding randomization methods, allocation sequence generation, and blinding within the RCTs may lead to bias and can possibly distort the conclusions [51]. In our meta-analysis, allocation sequence generation was clearly described by the authors in two studies, while concealment and blinding of the patient, personnel, and observer were clearly reported by only one author. Power analysis calculations for minimum sample sizes were not performed in the study designs, and an adequate assessment of each outcome was reported by the authors in three trials only. In particular, universal definitions of postoperative biliary leak and common bile duct obstruction were not provided by all authors, and consequently a degree of author subjectivity may be a potential limitation in the assessment of complications. Moreover, no studies reported their intention-to-treat analysis for the outcomes. So while this systematic review and meta-analysis presents evidence to suggest that PDC is the optimal technique for dochotomy closure after LCBDE, properly powered and well-constructed RCTs are required to validate the preliminary results presented in this study.

In summary, this comprehensive meta-analysis suggests that PDC after LCBDE is as safe as TTD. Furthermore, PDC is associated with a lower rate of postoperative biliary peritonitis when compared with TTD, and also shorter operating times, postoperative hospital stay, and a reduction in median hospital expenses. These results suggest that primary bile duct closure may be the optimal strategy for dochotomy closure after LCBDE. However, large well-designed and adequately powered RCTs that compare primary duct closure and T-tube insertion, as well as primary duct closure with and without biliary stenting are still required to validate these observations and also to give further clarity on whether there should be selective internal drainage of the biliary system in certain circumstances.