Safe Cholecystectomy Multi-Society Practice Guideline and State of the Art Consensus Conference on Prevention of Bile Duct Injury During Cholecystectomy

L. Michael Brunt, MD; Daniel J. Deziel, MD; Dana A. Telem, MD, MPH; Steven M. Strasberg, MD; Rajesh Aggarwal, MD; Horacio Asbun, MD; Jaap Bonjer, MD; Marian McDonald, MD; Adnan Alseidi, MD; Mike Ujiki, MD; Taylor S. Riall, MD, PhD; Chet Hammill, MD; Carol-Anne Moulton, MD; Philip H. Pucher, MD; Rowan W. Parks, MD; Mohammed T. Ansari, MD, MMedSc, MPhil; Saxon Connor, MD; Rebecca C. Dirks, MD; Blaire Anderson, MD; Maria S. Altieri, MD; Levan Tsamalaidze, MD; Dimitrios Stefanidis, MD, PhD


Annals of Surgery. 2020;272(1):3-23. 

In This Article

Recommendations for Future Studies

Regarding the Need to Grade Severity of AC and History of Prior Attacks of AC

Studies that examine the relationship between BDI and AC should match patients at baseline both for severity grade of AC and history of prior attacks of AC.

Method of Diagnosis of AC

The diagnosis of AC should be documented in future studies following well-accepted clinical criteria such as TG18 diagnostic criteria, histologic findings of acute inflammation, or both. If documentation of AC is based on diagnostic codes such as ICD codes, investigators should ensure that the diagnostic codes conform with the preceding criteria.

Regarding Classification of Timing of Surgery in Studies of AC

In AC, for the purposes of reporting standardization and ability to compare results among studies, we suggest that the interval between onset of symptoms and time of operation should be defined in 4 phases:

Phase 1: Onset of symptoms to 72 hours. Inflammation expected to be favorable for cholecystectomy. Tissue swelling due to edema.

Phase 2: 72 hours to 10 days. Inflammation expected to be less favorable for cholecystectomy. Due to tissue swelling and increased vascularity.

Phase 3: 10 days to 6 weeks. Inflammation expected to be much less favorable for cholecystectomy. Acute and chronic inflammation.

Phase 4: 6 weeks or later. Inflammation expected to be more favorable again for cholecystectomy. Predominately chronic inflammation.

There is also some justification for a period greater than 12 weeks but there is very little information in the literature that this time period had been the subject of study.

Question 9: Should STC versus total laparoscopic or open cholecystectomy be used for mitigating the risk of BDI in marked acute inflammation or chronic biliary inflammatory fusion?

Recommendation: When marked acute local inflammation or chronic cholecystitis with biliary inflammatory fusion of tissues/tissue contraction is encountered during LC that prevent the safe identification of the CD and artery, we suggest that surgeons perform STC either laparoscopically or open depending on their skill set and comfort with the procedure (expert opinion).

Summary of Evidence

No direct comparative evidence was found addressing this question. Indirect qualitative comparisons from case series of STC versus total cholecystectomy was considered critically flawed because of nonexchangeability of surgical populations due to confounding by indication for subtotal versus total cholecystectomy (see also discussion Question 2).

Narrative Synthesis

Only 1 article directly compared STC to LC with BDI as an outcome metric.[86] This retrospective study used administrative data from the University HealthSystem Consortium Database to compare LC to STC. Patients who underwent laparoscopic STC were older (56 vs 48 years), more likely male (54.2% vs 32.3%), and had higher severity of illness scores on admission (9.2% vs 3.5%) (P < 0.001) compared to standard LC. On univariate analysis, patients who underwent STC experienced longer lengths of stay (4 vs 3 days), higher readmission rates (11.9% vs 7.0%) and higher mortality (0.82% vs 0.28%) (P < 0.05). Following a 1:1 propensity score match, no differences in outcomes were demonstrated between patient groups. Thus, the authors concluded that STC is a safe and feasible alternative to LC in well-selected patients. The strength of this conclusion is dampened by the retrospective nature of the study and confounding variables such as intraoperative details, surgeon factors and patient factors which were not accounted for.

Of the remaining articles assessed, none directly compared outcomes and focused on the safety and feasibility of each technique. A systematic review identified 30 studies of 1200 patients who underwent STC.[71] BDI was the primary outcome measure with other complications as secondary outcomes. The approach was laparoscopic in 72.9%, open in 19.0%, and laparoscopic converted to open in 8.0%. Reasons for cholecystectomy were AC in 72%, cirrhosis/portal hypertension in 18.2%, gangrene or perforation in 6.1%, and Mirizzi syndrome in 3%. BDI was reported in 0.08%, bile leak in 18.0%, subhepatic collections in 2.9%, retained stones in 3.1% of patients, need for post-op ERCP in 4.1%, and re-operation in 1.8%. The mortality rate was 0.4%. When outcomes of laparoscopic versus open STC were compared, the laparoscopic approach was associated with less risk of sub-hepatic collection (OR 0.4, 95% CI 0.2–0.9), retained stones (OR 0.5, 95% CI 0.3–0.9), wound infection (OR 0.07, 95% CI 0.04–0.2), reoperation (OR 0.5 95% CI 0.3–0.9), and mortality (OR 0.2, 95% CI 0.05–0.9), but had more bile leaks (OR 5.3, 95% CI 3.9–7.2). Bile leaks resolved spontaneously in 31.2%. Of note, there was no standardization of technique across the various studies for how STC was performed and outcomes were not compared to total cholecystectomy.


Because no admissible direct or indirect evidence addressed this guideline question, the panel deemed that a STC is likely to limit the risk for BDI when operative conditions prevent clear anatomic identification by avoiding dissection in the hepatocystic triangle. Both open and laparoscopic approaches for STC were considered relevant based on patient and surgeon factors.

Question 10: Should standard 4-port lap cholecystectomy versus reduced port LC (single incision laparoscopic cholecystectomy, SILC) versus robotic cholecystectomy versus open cholecystectomy versus other techniques be used for limiting the risk or severity of BDI in candidates for cholecystectomy?

Recommendation: For patients requiring cholecystectomy, we suggest using a multi-port laparoscopic technique instead of single port/single incision technique (conditional recommendation, moderate certainty of evidence).

Summary of Evidence

No direct comparative evidence addressed this outcome. Indirect comparison was made from evidence from 2 systematic reviews (including total >507,918 patients) of single-arm cohort studies of single-port versus standard port approach. The pooled effect estimate for BDI was 0.72% (simple average of the data across studies) versus 0.32%–0.52% (pooled range). The outcomes on which results favored single port LC were analgesic use, cosmesis, and quality of life. The magnitude of these effects was imprecise and clinically small to trivial. The undesirable effects for single port LC include BDI, total severe complication, operative time, port site hernia, and conversion to open procedure. The magnitude of these effects was conservatively judged moderate by the panel.

Narrative Synthesis

In a 2012 systematic review of BDI in SILC[139] (45 cohort studies, n = 2626), a pooled rate of BDI with SILC of 0.72% was found, potentially representing a higher rate of BDI than reported in previously reported large scale pooled data for 4-port LC (0.32%–0.52%).[11] A 2018 meta-analysis of 24 randomized trials (n = 4518) comparing SILC to 4-port LC reported a significantly increased incidence of major complications (Clavien-Dindo grade III or greater) in SILC, 2.7% versus 1.1%, RR 2.02 (95% CI 1.29–3.15).[140] The quality of evidence in this review was moderate with good consistency of results.

Other outcomes were assessed by 3 smaller reviews, which were deemed of low or very low quality with high bias risk. Postoperative analgesic use was compared for SILC and 4-port LC in a 2013 review of 2 cohort studies, 1 RCT, n = 361). No difference was reported in postoperative opioid use, weighted mean difference −3.78 mg (−13.78 to 6.22 mg).[141] Conversion rates were compared for SILC and 4-port LC in a 2014 review (27 RCTs, n = 2049), with no difference between groups (0.2% in both groups).[142] A 2014 Cochrane review of 9 RCTs of 4-port versus reduced port LC (7 trials assessed SILC; 2 trials 3-port LC) reported longer operative time for reduced-port LC (mean difference +14.4 95% CI 6.0–23.0 minutes).[143]


Although for individual outcomes, the certainty ratings ranged from very low to moderate, the evidence on critical outcomes (BDI, severe complication) consistently favored standard port LC with certainty rating for severe complications judged to be moderate. As such, the highest certainty evidence informed overall certainty. Although there may be variability in how much patients value cosmesis, the panel believes that almost all patients will value the remainder of outcomes greater than cosmesis alone. Although evidence favors standard port, the panel acknowledged that in highly experienced hands, single port may yield similar outcomes as standard port.

Question 11: Should interval LC versus no additional treatment be used for patients previously treated by cholecystostomy drainage?

Recommendation: In patients with acute calculous cholecystitis previously treated by cholecystostomy who are good surgical candidates, we suggest that interval cholecystectomy is preferred after the inflammation has subsided. For poor or borderline operative candidates, we suggest a nonsurgical approach that may include percutaneous stone clearance through the tube tract or tube removal and observation if the CD is patent (expert opinion).

Summary of Evidence

No direct comparative evidence was found that addressed this question. Making indirect comparisons from case series of patients treated by cholecystostomy who had no additional treatment versus patients who had interval cholecystectomy was considered flawed because of nonexchangeability of the populations due to confounding in the predicted operative risks of the patients. Stated otherwise, in studies of this type, the group of patients who do not undergo interval cholecystectomy will be composed of patients who are candidates for surgery and patients who are not candidates for surgery in unknown proportions. Because such a group contains patients who are not candidates for cholecystectomy, they cannot be directly compared to a group of patients all of whom were considered to be candidates for surgery and underwent cholecystectomy. We found no studies in which this issue was addressed by propensity scoring or randomization of fit candidates for surgery to cholecystectomy or observation.

Narrative Synthesis

A systematic review on the use of cholecystostomy as a treatment for AC performed in 2007 identified 53 studies that included 1918 patients.[144] The short-term mortality after cholecystostomy tube placement, 30-day or in-hospital, depending on which was reported in the manuscripts, was 15.4%. Thirty-eight percent of patients had an interval elective cholecystectomy and 4.5% required an emergent cholecystectomy. The operative mortality of patients undergoing elective interval cholecystectomy was 0.96% and mortality was 13% in patients undergoing emergent cholecystectomy. Emergent cholecystectomy included cholecystectomy during the index admission due to therapeutic failure and procedural complications in addition to patients experiencing recurrent cholecystitis after removal of the cholecystostomy tube.

A retrospective cohort study published in 2013, using a population-based administrative database from Canada, identified 890 patients who had a cholecystostomy tube placed for AC from 2004 to 2011.[145] In-hospital mortality was 5%. At 3 months, 25% of patients had a cholecystectomy and 10% had died without undergoing cholecystectomy. At 1 year, 40% had cholecystectomy and 18% had died. Thirty-day and 1-year postoperative mortality in the cholecystectomy group was 2% and 6%, respectively. In the 866 cholecystostomy patients discharged without cholecystectomy, the risk of gallstone-related emergency department visit or hospitalization was 23% at 3 months and 49% at 1 year. In-hospital mortality of gallstone-related admissions was 1%.

A single-center retrospective review from the United States identified 288 patients who had a cholecystostomy tube placed for acute calculous cholecystitis from 1997 to 2015.[146] In-hospital mortality was 9%. Tube dysfunction occurred in 46% of patients with 28% requiring re-intervention. Thirty-six percent of patients underwent subsequent elective cholecystectomy, and the rate of recurrent biliary events was 6.8% versus 21.1% in patients who did not have a cholecystectomy (5.8% vs 18.5% at 1 year). The risk of biliary-related deaths in the patients who did not have a cholecystectomy was 5%. A single-center retrospective review from Denmark identified 278 patients who had a cholecystostomy tube placed for acute calculous cholecystitis from 2002 to 2012.[147] Thirty-day mortality was 4.7%. Of 234 patients discharged, 23.5% were readmitted for recurrent cholecystitis, 54.7% were followed for a median of 5 years without recurrence, and 21.8% had an elective interval cholecystectomy. In another single-center retrospective review from the United States, 245 patients between 2009 and 2012 had a cholecystostomy tube placed.[148] Of the 202 patients who survived to discharge, 48% had calculous cholecystitis and 41% had acalculous cholecystitis. Seventy-one patients (35%) were determined to be operative candidates, the majority of whom had calculous cholecystitis and had a planned interval cholecystectomy. Two smaller series of 93 and 68 patients reported recurrent gallbladder disease in 19% and 41%, respectively of patients who did not undergo planned interval cholecystectomy.[149,150]

In addition to a patient's surgical risk, the subgroups of patients with calculous and acalculous cholecystitis need to be considered. Some data in the literature demonstrates a lower risk of recurrent biliary symptoms after cholecystostomy tube removal in patients with acalculous cholecystitis. The largest study, with 88 patients, demonstrated a 2.7% risk of recurrence over an 8-year follow-up.[151] Smaller studies have demonstrated a 7%–14% risk of recurrence.[152–154]

It should be noted that an important issue related to management of patients with a cholecystostomy tube is the indications for placement of a cholecystostomy tube. CHOCOLATE, a Dutch randomized trial that compared LC to percutaneous cholecystostomy in high risk patients with acute calculous cholecystitis was recently published.[155] The results from this study strongly favor LC over cholecystostomy tube placement.


In the available case series, the desirable anticipated effect of interval cholecystectomy is avoidance of recurrent gallbladder-related symptoms which occurred in 20%–50% of patients who did not undergo interval cholecystectomy. There is also some evidence that in this group recurrent symptoms were associated with urgent cholecystectomy, which has a higher rate of open cholecystectomy and postoperative complications, whereas patients who undergo elective interval cholecystectomy are more likely to have the cholecystectomy completed laparoscopically with its associated benefits. The undesirable anticipated effect of interval cholecystectomy is increased cholecystectomy-related complications including mortality. In the largest case series, the 30-day mortality after interval cholecystectomy was 2%. However, this will vary significantly based on criteria used to select patients for elective interval cholecystectomy. The available data is retrospective and it is likely that the patients selected for elective interval cholecystectomy were chosen for their perceived ability to tolerate surgery. The guideline panel judged; therefore, that in the subgroup of patients determined to be fit for surgery, the balance of effects favors the intervention, and the anticipated desirable effects were determined to be moderately substantial. In the subgroup of patients determined to be unfit for surgery, the balance of effects probably favors the alternative management option.

Question 12: Should conversion of LC to open cholecystectomy versus no conversion be used for limiting the risk or severity of BDI during difficult LC?

No recommendation was made as the current evidence comparing conversion versus no conversion to open cholecystectomy to limit/avoid BDI in the difficult cholecystectomy is insufficient.

Thirteen studies were reviewed but none fit within the criteria for this analysis. No evidence was found for the outcome BDI or any of its proxy outcomes.

Recommendations for Future Studies: (1) We suggest the conduct of prospective and retrospective comparisons of clinical outcomes of various "bail-out" options for the difficult cholecystectomy that include conversion to open, STC, and procedure abandonment. (2) We further suggest the development and establishment of valid evidence for a "procedure difficulty score" for LC and study of its effectiveness in limiting BDI risk.

Question 13: Should surgeons take a time out to verify the CVS versus no time out be used for limiting the risk or severity of BDI during LC?

Recommendation: Current evidence is insufficient to make a recommendation. However, as a best practice, we suggest that during LC, surgeons conduct a momentary pause for the surgeon to confirm in his/her own mind that the criteria for the CVS have been attained before clipping or transecting ductal or arterial structures (expert opinion).

Of 2 studies reviewed, no evidence was found for the outcome BDI or any of its potential proxy outcomes.[156,157] Although no evidence exists to answer this question, the GDG panel felt based on experience that the incorporation of a momentary pause to verify the appropriateness of the CVS before any structures have the potential to decrease the risk of BDI. As the most common cause of BDI is a misperception of anatomy, the momentary pause is an opportunity to verify that what one is seeing is likely the correct anatomy. This practice should be easily implementable and without significant effort or delay required.

Recommendation for Future Studies: We suggest incorporation of a "critical view momentary pause" in all prospective studies of LC.

Question 14: Should 2 surgeons versus 1 surgeon be used for limiting the risk or severity of BDI during LC?

No recommendation was made as the current evidence comparing 2 versus 1 surgeons for limiting/avoiding BDI in cholecystectomy is insufficient.

A single study was identified from the literature search, but was excluded as it did not include evidence for the outcome BDI or any of its proxy outcomes.[158] Performing cholecystectomy with 2 surgeons present is not feasible in most settings. Given the potential beneficial effects of the presence of a second surgeon and the added experience that becomes available, the advice of a second surgeon is often very helpful under conditions in which the dissection is stalled, the anatomy is unclear, or under other conditions deemed "difficult" by the surgeon.

Retrospective assessment of case notes where the involvement of multiple surgeons was recorded should be possible from billing and electronic records, and linkage of these cases to outcomes could provide insight into the usefulness of 2-surgeon cholecystectomy. The effect of access to and/or involvement of subspecialist hepatobiliary surgeons and impact on outcomes should also be assessed. Prospective multi-center cohort studies are desirable to capture the effect of multiple surgeon involvement in clinical outcomes.

Question 15: Should CVS coaching of surgeons versus no specific CVS coaching be used for limiting the risk or severity of BDI during LC?

Recommendation: We suggest as a best practice continued education of surgeons regarding the CVS during LC that may include coaching (conditional recommendation, very low certainty of evidence).

Summary of Evidence

Five studies were identified from the literature search for review. One study directly addressed this question but was not considered appropriate for inclusion due to numerous flaws (underpowered to detect BDIs, inclusion of bile leaks as BDIs, confounding by surgeon skill).[159] Therefore, One before and after study which addressed the proxy outcome of quality of CVS was included. In a group of 5 practicing surgeons who received CVS coaching, their scores on the Strasberg scale improved significantly from 1.75 at baseline to 3.75 after training (very low certainty evidence because of study design and fragility of effect).[160]

The GDG agreed that there was evidence from other surgical domains in support of this judgment. A systematic review of quality of evidence for surgical coaching and its impact on outcomes reported a positive impact on surgical performance, with strong evidence particularly in the domain of technical skills.[161] A 2015 randomized trial reported significantly improved global and technical performance in porcine LC when comparing coaching to no coaching.[162]


Obtaining a high-quality "critical view" was considered paramount by the GDG to avoid BDI during LC. Given that current evidence suggests that the majority of surgeons do not routinely obtain the CVS,[160] and the benefits of education and coaching in improving performance and changing behavior demonstrated in several fields including surgery,[161] the GDG felt that coaching should be employed to improve the quality of dissection and of the CVS and may lead to decreased risk of BDI.

Question 16: Should training of surgeons by simulation methods or video-based education versus alternative surgeon training be used for limiting the risk or severity of BDI during LC?

No recommendation was made as the current evidence comparing simulation or video-based training versus alternative surgeon training modalities on limiting/avoiding BDI during LC is insufficient.

Five studies were identified from the initial literature search.[163–167] No evidence was found for the outcome BDI or any of its proxy outcomes.

Recommendation for Future Studies: We suggest the conduct of prospective large-scale multi-center studies to determine the role of simulation versus video-based versus alternative surgeon training modalities on limiting/avoiding BDI. More realistic simulators incorporating immersive virtual technology and advanced haptics should be considered for development, and their effectiveness studied.

Question 17: Should more surgeon experience versus less surgeon experience be used for mitigating the risk BDI associated with LC?

Recommendation: We suggest that surgeons have a low threshold for calling for help from another surgeon when practical in difficult cases or when there is uncertain of anatomy (conditional recommendation, very low certainty of evidence).

Summary of Evidence

From 23 studies identified from the initial literature search, a single 2014 observational study was included which examined 53,632 LC's from a US billing database and linked surgeon codes to the Fundamentals of Laparoscopic Surgery database.[7] The authors reported that more experienced surgeons (mean 20.7 vs 6.1 years in practice) experienced a lower rate of BDI [0.14% vs 0.47%, P = 0.0013, RR 0.27 (95% CI 0.13–0.57)] compared to less experienced surgeons.


Despite the large effect size and effectively nonexistent undesirable effects, a strong recommendation was not provided due to acceptability and feasibility considerations, especially in smaller community hospitals or rural areas (where only one surgeon may be available). The GDG felt that senior support could be sought for difficult cases identified pre- or intraoperatively, or that referral to more experienced centers should be considered where such support is not available.

Recommendation for Future Studies: We suggest the conduct of prospective research studies to develop evidence-based guidelines and measure the impact of interventions (such as the SAGES Safe Cholecystectomy initiative) for physicians who are in transition in practice/from residency/fellowship to independent practice, to mitigate the risk of BDI associated with LC.

Additional studies to measure the impact of surgeon experience on outcomes in LC are needed. Prospective comparisons of the effectiveness of the various BDI mitigating training methods on the LC outcomes of early career surgeons should be considered.

Question 18: For patients with BDI during LC (in the OR or early postoperative period), should the patient be referred to a specialist with experience in biliary reconstruction or should the reconstruction be performed by the operating surgeon?

Recommendation: When a BDI has occurred or is highly suspected at the time of cholecystectomy or in the postoperative period, we recommend that surgeons refer the patient promptly to a surgeon with experience in the management of BDI in an institution with a hepato-biliary disease multispecialty team. When not feasible to do so in a timely manner, prompt consultation with a surgeon experienced in the management of BDI should be considered (strong recommendation, low certainty of evidence).

Summary of Evidence

No randomized control trials or systematic reviews were found that addressed this question. There were 3 retrospective studies and 44 case series, the majority of which include only patients with BDI repaired at expert centers. Studies of BDI series that compared outcomes of repair by the primary surgeon/institution versus HPB centers were considered flawed because they consisted mostly of failed primary repairs and lack a denominator of total repairs by the primary surgeon. Seven observational studies summarized below that included a total of 1392 patients with BDI addressed this question and showed a large consistent effect across a number of outcomes (rates of cholangitis, bile leak, stricture, reoperation, other intervention, and death) that favored specialist surgeons over primary surgeons. The studies were judged to be at high risk of bias due to unclear comparability and an unknown degree of missing data. An umbrella systematic review of 32 systematic reviews covering 15 surgical procedures provided indirect evidence additionally favoring higher surgeon experience or volume (R-AMSTAR score 33/44).[168]

Narrative Synthesis

  1. Perera et al[169] presented a series of 200 patients treated for major BDI with a median follow-up of 60 months. In 52% of the cases the anatomy was described as "straightforward" during the LC; 72% of the injuries were a major type E injury and 13% a type D. Of 157 patients managed surgically, 112 (71%) underwent reconstruction by specialist HPB surgeons and 45 patients (29%) were operated by non-HPB surgeons before referral. Multivariate analysis showed that on table repair done by nonspecialist surgeons was an independent risk factor for recurrent cholangitis (33% vs 11%), biliary strictures (69% vs 17%), redo reconstructions (53% vs 3%), and overall morbidity (82% vs 25%), P < 0.001.

  2. Stewart et al[170] reported outcomes in 137 patients with BDI repaired by HPB surgeons after no prior attempt at repair compared to 163 patients referred after repairs by the primary surgeon. The success rate of repair by HPB surgeons was 90.5% versus 20.9% with primary surgeon repair.

  3. Xu et al[171] examined outcomes of repair of 77 patients with BDI across 15 provinces in China. The success rate of repair by HPB specialists in 42 patients was 83.3% compared to a success rate of 31.4% in 35 patients repaired by the primary surgeon.

  4. DeReuver[172] in a study of 500 BDIs referred found on multivariate analysis that factors associated with failure included secondary referral after prior surgical, endoscopic, or radiologic interventions.

  5. Sicklick et al[173] reported a single-institution retrospective analysis on 200 patients treated for major BDI. Eighty-one patients (44%) had attempted repairs at outside institutions before referral. Only 15 of 89 repairs (17%) required no further reoperation. In 31 patients (34.8%), the original laparoscopic surgeon performed an end-to-end ductal anastomosis and in 18 patients (22.5%) the surgeon was unable to repair the injury. Definitive repair at the specialty center with a Roux-en-Y hepatico-jejunostomy was carried out in 98% of patients. Complications occurred in 42.9% of patients but no patient required reoperation.

  6. Thomson et al[174] reported on 123 BDI, 87 (70.7%) occurring during LC and 33 (26.8%) during open cholecystectomy. Of 55 patients who had an attempted repair before referral, only 12 (22%) required no further surgical intervention. Of the 42 patients who required further surgery, 27 had a failed primary ductal repair at the outside institution. A successful outcome of repair was achieved by a team experienced in the management of BDI in 89% of cases.

  7. Rystedt et al[175] reported results of early repair in 140 of 155 BDI detected during LC, 90 of which were initially repaired by the index surgeon. The most common repair was suture over T-tube and only 17% of patients had a Roux-en-Y hepatico-jejunostomy. In this study, 59% of the injuries were Hannover Grade C1 (<5 mm lesion). Strictures were reported at 6 months in 18% of patients.

Notably, 2 studies found that outreach programs in which an experienced HPB surgeon travels to the primary center can also achieve good outcomes.[174,176]

Together these studies suggest that early referral to a tertiary center with experienced hepatobiliary surgeons and associated interventional radiologic and endoscopic services seems necessary to ensure optimal results. In addition, abundant indirect evidence from other surgical domains has previously established a positive relationship between surgeon volume and surgical outcomes. Quantitative, pooled data from an umbrella systematic review covering 15 surgical procedures that included abdominal aortic aneurysm, colorectal cancer, esophagectomy, pancreatic cancer resection, and other procedures favored higher surgeon experience or volume.[168] Nonpooled results from individual studies on the Norwood procedure, trauma, bariatric surgery, radical prostatectomy, total knee arthroplasty, and coronary artery bypass also supported the same relationship between improved outcomes and surgeon volume.


In summary, there is a strong clinical rationale and both direct and indirect evidence favoring specialty repair despite the very low certainty of direct evidence addressing this question. Consequently, the GDG invoked 1 of the 5 paradigmatic situations for a strong recommendation originating in this setting (potential for catastrophic harm).[34] The complexity in assessing the extent of BDI and potential associated vascular injury, and the type of surgery entailed in the repair, are significantly different than for LC. Therefore, high volume experience in LC cannot be generalized to outcomes of repairs of BDI. No concerns were noted by the panel regarding the generalizability of the systematic review evidence. Further, the undesirable effects secondary to a potential delay related to a specialist referral were considered small or trivial, contingent to preparing the patient well for such a referral/transfer, that is, placement of drains by primary surgeon. As such, the balance of benefit and harms were judged to strongly favor the intervention. Consequences of a poorly repaired or failed repair of BDI include catastrophic harms such as cholangitis, bile leak, biliary stricture, sepsis, need for reoperation or other interventions, and liver failure, all of which may potentially lead to death.

Recommendations for Future Studies: Implementation pathways should be developed with creation of regional fast tract BDI referral pathways to offer advice and contribute to immediate treatment strategies when a BDI due to LC occurs.

Additional Panel Recommendation: We suggest the development of national quality improvement initiatives for the prevention of BDI after cholecystectomy. The initiative(s) should be capable of identifying and characterizing BDI in the population under study.

One of the challenges of efforts to impact the rate of BDI is that it is extremely difficult to define the actual incidence and whether it might change over time. To determine this, large numbers of cholecystectomies need to be evaluated, and the administrative data that is currently used for this purpose from payors or other groups has inherent limitations, inaccuracies, and significant risk of bias. BDI is not just a USA or North American phenomenon, it is a worldwide phenomenon, which is why quality improvement initiatives should be at the national level whether within a country, a region, or within a professional surgical organization or society. The collective weight of the societies participating in this conference should work toward implementation strategies to move this recommendation forward.