On-table hepatopancreatobiliary surgical consults for difficult cholecystectomies: A 7-year audit

Kai Siang Chan, Elizabeth Hwang, Jee Keem Low, Sameer P Junnarkar,Cheong Wei Terence Huey, Vishal G Shelat

Department of General Surgery, Tan Tock Seng Hospital, 308433, Singapore

Keywords:Bile duct injury Cholecystectomy Cholecystitis On-table consult Quality indicators Hepatopancreatobiliary surgery

ABSTRACT

Introduction

Subspecialty training has reduced the “generality” of general surgery as "organ-specific" pathologies are often referred to their respective subspecialty teams. Despite advances in training standards and specialization or fellowship opportunities, subspecialty training and trained experts are not routinely available in all healthcare settings. Even where trained experts are available, the surgical disease may be so common in the community that a handful of experts are not adequate to cater to communities’ healthcare needs, and thus, a general surgeon is expected to manage the diseased organ. Cholecystectomy is one such surgery where licensing does not require special training or accreditation. Globally, cholecystectomy is considered a general surgical operation and performed by surgeons with diverse subspecialty interests.

Hepatopancreatobiliary (HPB) surgery is a relatively new subspecialty that has gained recognition over the past two decades [1] .HPB referral of all patients with gallbladder disorders who require cholecystectomy is unwarranted and may lead to unnecessary delays in care for patients who require subspecialty care, such as hepatic or pancreatic malignancies. Although general surgeons are trained to perform cholecystectomy; they are not trained or experienced in managing all the cholecystectomy complications, especially bile duct injury (BDI). Hence, patients should be referred to HPB surgeons when an intraoperative difficulty arises, or major postoperative complications occur. It is shown that delay in referrals results in increased morbidity [ 2 , 3 ]. Thus, an on-table HPB consult is sought under challenging situations, such as anatomical variations, the severity of the lesion, or intraoperative complications.

Involvement of HPB specialists may alleviate the morbidity, and possibly improve clinical outcomes. Several centers have staffing and resources available to provide an on-table consult “service”. A study by Silva et al. reporting on the on-table repair of BDIs by HPB specialists in 22 patients demonstrated the safety and feasibility of the on-table consult “service” [4] . However, service availability does not equate to accessibility, and the onus remains on the primary surgeon to call for help. Jin et al. reviewed the literature on cognitive psychology and concluded that reputation and ego pressures might interfere with the thought processes needed to execute the tasks at hand [5] . Some surgeons may prefer to rely on their abilities and have a high threshold before escalating to the relevant specialties. A surgeon also needs to know his or her abilities and limitations, including but not limited to, technical skills,teamwork, and communication skills. Hence, despite the availability of an on-table expert HPB consult, delays are likely and adversely affect patient outcomes. Though “call for help” is accepted as an established standard of care in difficult situations, what happens next remains unknown. To our knowledge, there are no studies that document the outcomes of patients who had on-table HPB consults during difficult cholecystectomy. This study aimed to investigate the outcomes of patients with an on-table HPB consult at a tertiary care institution where general surgical teams routinely perform cholecystectomy.

Methods

This is an audit of 87 patients who required an on-table HPB consult from 2011 to 2017. Patients who required HPB consult for oncologic clearance (n= 21), non-oncological multi-visceral involvement such as trauma (n= 6), hernia repairs (n= 4), and others (n= 6) were excluded. On-table consult requests in the remaining 50 patients were primarily made for difficulty during cholecystectomy. No approval from the institutional review board was required for this audit, which was meant for local HPB referral service quality measures. Patient identifiers were not collected nor stored, and patient contact was not made for this audit. No attempts were made by the study team to access the medical records of patients via the national electronic health record system. The conduct of this study was in accordance with the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE)statement for retrospective studies [6] .

Definition of on-table consult

For this study, we defined the type of consults as "proactive"and "reactive". A proactive consult was before surgical incision, and a reactive consult was after the surgical incision. Proactive consult entailed the primary surgeon anticipating difficulty and preemptively alerting HPB service before surgery, and all other categories of consults were considered reactive. With regards to cholecystectomy, the reactive consult was grouped into three categories: (1)anatomical, unclear anatomy or anatomical variations; (2) pathological, the presence of dense adhesions, the severity of underlying disease complicating the surgery, such as gangrenous or emphysematous cholecystitis, impacted stones and others; (3) surgical, intraoperative complications such as bile leak, or structural injuries (BDI, cystic duct injury or bleeding from an artery or surgical bed). Operative records were reviewed from individual HPB surgeon logbooks (Low JK, Junnarkar SP, Huey CWT, and Shelat VG),and the type of consult request was retrospectively determined by two authors (Hwang E and Shelat VG). BDI was classified according to the Strasberg classification [7] . Readmission was defined as any re-admission to the hospital for a related disease within 30 days.Morbidity was defined as any intraoperative or postoperative complications. 30-day and 90-day mortalities were defined as all-cause deaths within 30 days and 90 days after surgery.

Treatment protocol

Patients who underwent cholecystectomy for asymptomatic gallstones or biliary colic were admitted as day surgery cases(23 h hospital stay) for elective cholecystectomy by general surgical teams. Patients who presented with acute cholecystitis were admitted and were similarly managed by general surgical teams.Local algorithm for managing patients with acute cholecystitis involved at least one set of blood cultures before administering empirical parenteral amoxicillin-clavulanic acid with a stat dose of gentamicin 3–5 mg/kg body weight. The timing of surgical management was according to the Tokyo Guidelines 2013 for acute cholecystitis [8] . Index admission cholecystectomy is widely practiced locally by all general surgery teams. Local HPB service adopts a universal cholecystectomy policy and provides cover for emergencies, including consults for difficult cholecystectomies [9] . Universal cholecystectomy policy entails index admission cholecystectomy for all indications (e.g., biliary pancreatitis) regardless of onset of abdominal pain. Patients who were deemed unfit for cholecystectomy, who refused surgery, or who could not be operated on due to higher bleeding risks (ongoing non-aspirin antiplatelet medications or anti-coagulants) were offered interval cholecystectomy after risk stratification and optimization. Interval cholecystectomy was offered both as day surgery or inpatient admission for co-morbidity management. Deviation in postoperative recovery was investigated according to clinical judgment. Image-guided postoperative percutaneous pigtail drain was inserted for patients with intra-abdominal collection. Management of BDIs was dependent on the extent of BDI according to the Strasberg classification as well as surgeon experience and judgment [ 7 , 10 ].

Statistical analysis

The data extracted were tabulated into an excel sheet and transposed into SPSS version 25 (SPSS inc., Chicago, IL, USA) for statistical analysis. Mean imputation was performed for missing data values where＜10% was missing. Shapiro-Wilk test of normality was performed for all continuous variables and revealed a non-parametric distribution. Categorical variables and continuous variables were described as percentages and median [interquartile range (IQR)], respectively. This study’s statistical review was performed by one of the co-authors qualified in biomedical statistics(Shelat VG).

Results

Fifty patients with a median age of 62.5 years (IQR 50.8–71.3 years) required on-table HPB consult for cholecystectomy during the study period. Thirty-three (66%) patients were male, and eight(16%) patients had underlying HPB co-morbidity. Patient demographics and clinical profiles are summarized in Table 1 . Gallbladder wall was thickened in all patients (median 5 mm, IQR 4–7 mm), and common bile duct was of normal caliber in all patients(median 5 mm, IQR 4–6 mm).

Table 2 summarizes the perioperative details of patients. Median length of operation was 165 min (IQR 124–209 min), medianblood loss was 100 mL (IQR 50–200 mL), and median length of hospital stay was five days (IQR 3–7 days). Morbidity and 30-day readmission were 22% (11/50) and 6% (3/50), respectively. There was no 90-day mortality. Most of patients were initially managed by laparoscopic approach (48/50, 96%), 15 (31%) required an open conversion. Majority of the conversions (9/15, 60%) were initiated by non-HPB surgeons prior to seeking on-table HPB consult. Reasons for conversion to an open approach were dense adhesions,unclear anatomy and/or shrunken or contracted gallbladder (n= 6,40%), BDI (n= 3, 20%), anatomical variations (n= 2, 13%), bile leak(n= 2, 13%), and bleeding from cystic artery or right hepatic artery(n= 2, 13%).

Table 1 Clinical profile of patients who underwent cholecystectomy with on-table hepatopancreatobiliary (HPB) consults.

Table 2 Perioperative details of patients who underwent cholecystectomy with on-table hepatopancreatobiliary (HPB) consults.

Most of the referrals were reactive (49/50, 98%). One patient had a proactive referral to confirm arterial anatomy for anatomical variation.

Reasons for on-table HPB consults are summarized in Table 3 .The most common reason for on-table HPB consults was the presence of unclear anatomy (15/50, 30%), such as difficulty identifying Calot’s triangle or the presence of anatomical variations as the short cystic duct. Nine (18%) patients were referred for the presence of dense adhesions and/or contracted gallbladder causing difficulty in dissection. Eight (16%) patients were referred due to the presence of impacted stones in Hartmann’s pouch. Bile leak was present in three patients (6%); intraoperative cholangiogram was performed for all, and none had BDI. Three (6%) referrals were for bleeding: one each from the cystic artery, right hepatic artery, and gallbladder bed, respectively. Three patients (6%) each had consults for cystic duct injury and BDI, respectively. HPB specialists managed to control cystic duct stump in all three patients, and BDI is discussed separately below.

BDI

Three patients were referred for BDI. All of them underwent conversion to open cholecystectomy and immediate repair. Patient A was a 73-year-old female who underwent an emergency laparoscopic cholecystectomy. Strasberg type D BDI was confirmed by intraoperative transcystic cholangiogram, and primary lateral wall repair of the common hepatic duct was performed. Her postoperative course was complicated by intra-abdominal collection. This was managed by image-guided percutaneous drainage and a course of intravenous antibiotics. She was discharged after two weeks of hospital stay and remained well at a six-month follow-up. Patient B was a 47-year-old male who underwent an elective laparoscopic cholecystectomy, which was converted to open due to the presence of dense adhesions and Strasberg type D BDI. He underwent intraoperative cholangiogram which confirmed the nature of injury and primary lateral wall common hepatic duct repair over a T-tube was done. He recovered well postoperatively with no complications; however, he had readmission within 30 days for abdominal pain. Serum biochemistry showed normal liver function, and imaging did not show a dilated duct or intra-abdominal collection.He was managed expectantly, and T-tube was removed after seven weeks. Patient C was a 63-year-old female who underwent elective laparoscopic cholecystectomy. Due to intraoperative difficultyand obliteration of Calot’s triangle, a fundus-first cholecystectomy was performed instead. Intraoperative Strasberg type E1 BDI was sustained, confirmed by intraoperative cholangiogram. A Roux-en-Y hepaticojejunostomy was done, and postoperative recovery was uneventful.

Table 3 Reasons for patients who had on-table hepatopancreatobiliary (HPB) consult.

Discussion

To the best of our knowledge, this is the first report of on-table HPB consults for difficult cholecystectomy. Our study demonstrated that on-table HPB consults for difficult cholecystectomy showed 22% morbidity, 31% open conversion, and 6% BDI.

Cholecystectomy is a “bread and butter” procedure and commonly performed globally by general surgeons. Due to the high prevalence of gallstones in the community and structured residency training programs with fellowship opportunities, most general surgeons are comfortable in handling difficult situations. Thus,on-table HPB consults are uncommon, and the true incidence of such consults is unknown and there is paucity in literature about the outcomes of cholecystectomy patients where an on-table HPB consult was sought. Locally about two-thirds of cholecystectomies are done by non-HPB teams with an estimated annual caseload of about 500 cholecystectomies, and about 2.1% of patients are referred for on-table HPB consult at our institution [11] .

Safe cholecystectomy entails obtaining the critical view of safety (CVS) by careful dissection of Calot’s triangle. Anatomical difficulties and severe pathology can lead to scarring and fibrosis of the Calot’s triangle, obscuring the CVS, and thus, alternative bail-out strategies are essential skills for any general surgeon performing cholecystectomy. When faced with difficulty, the primary surgeon must decide on the next best course of action, considering patient safety as a primary interest. Bail-out procedures such as fundus-first cholecystectomy, open conversion, subtotal cholecystectomy, or cholecystostomy are potential options. In a local study on outcomes of 168 patients treated by subtotal cholecystectomy,Tay et al. has demonstrated the safety of subtotal cholecystectomy and reported no BDI, reoperation, nor 30-day mortality [11] . Besides bail-out strategies, abandoning the surgery or "calling for help" by seeking an on-table HPB specialty consult should be considered too [12] . “When in doubt, seek help” is a familiar dictum widely taught in surgical curricula and is considered good practice. However, autonomy, confidence, bravado, and feeling of embarrassment may cloud judgment and render "calling for help" difficult [5] . Existing guidelines do not outline the type of bail-out procedures surgeons should adopt; the choice depends on the surgeon’s clinical context, experience, and clinical judgment [13] . "Call for help" may be the second- or third-line approach a surgeon may have after exhaustion of other means. Surgeons may consider bailout procedures such as subtotal cholecystectomy or fundus-first cholecystectomy before escalation.

BDI is the Achilles’ heel of cholecystectomy and key performance indicator of safe cholecystectomy. BDI is the most severe complication of cholecystectomy with an incidence of 0.2%–0.6%,which impacts short-term outcomes, and diminishes the long-term quality of life [14] . Therefore, it is prudent to exhaust means to prevent BDI. Asking for an on-table HPB consult under challenging situations is integral to safe cholecystectomy [13] . The timing of referral to HPB surgeons is a topic of concern for BDIs. Several studies have shown that delay in referral for BDIs is associated with increased morbidity and a more extended recovery period [ 2 , 3 ]. There is no excuse for a general surgeon not to refer a patient with suspected or confirmed BDI to HPB service. Locally,on-table consults are easier to obtain with planned monthly rosters that provide round-the-clock HPB specialty coverage. In our study, all three patients with BDI were recognized intraoperatively,and HPB consults were sought. A large international Delphi consensus, including 372 respondents on BDI, concluded that advice from a second surgeon (not necessarily an HPB surgeon) might prevent misidentification of the common hepatic duct or common bile duct from the cyst duct by 18% [15] . One would expect an expert HPB consult to reduce the risk of misidentification or manage difficult cholecystectomies. We remain cautious to conclude high risk of BDI, due to the selection bias from included study population. This study involves a highly selected subgroup of patients with difficult cholecystectomy that triggered the primary surgeon to seek an HPB consult. In general, an HPB consult is only sought when primary surgeon fails to progress with cholecystectomy due to intraoperative difficulty or when BDI happens. It is an acceptable standard of care to call for HPB consult when BDI occurs, and our study endorses this culture of safety.This study does not prove that on-table consult reduces the injury or improves clinical outcomes, but reveals a real life scenario which can be encountered by any surgeon doing cholecystectomy procedure.

Dense adhesions, obliteration of Calot’s triangle, impacted stone in Hartmann’s pouch, and bile leak in the operative field are important milestones that should prompt a surgeon to pause and reflect. Inflammation was demonstrated to result in a three-fold increase in BDI in the study by Georgiades et al., and authors advocated prompt referral to HPB surgeons with open conversion to reduce the risk of iatrogenic BDI [16] . The comfort level of individual surgeons determines the threshold (the inflection point) to refer, and thus, we attempt to clarify the indication for proactive versus reactive consults. Proactive HPB consult may be considered for patients who are expected to have a complicated operation, provided that services are available and accessible. Male, age＞65 years, morbid obesity, leukocytosis (＞18 0 0 0/mm3), small contracted gallbladder, liver cirrhosis, cholecystoenteric fistula, history of abdominal surgery, and previous history of hepatobiliary diseases are some of the risk factors known to pose operative difficulty, and a high risk of open conversion and morbidity [ 11 , 13 , 17 , 18 ]. In our audit, the mean gallbladder wall thickness was 5 mm and suggested the chronicity of inflammation. Due to the retrospective nature of this audit, it is unclear if proactive HPB consultation would have prevented BDI. Nevertheless, surgeons face the dilemma of making consults in view of over-estimating the difficulty of the surgery and run the risk of making unwarranted referrals. However, surgeons should take into consideration of these risk factors and even if proactive referrals were not made, they should have a low threshold for escalating to an HPB specialist; early referral may improve outcomes [2] .

Our results show that locally an on-table HPB consult is always sought in patients with BDI. Several risk factors for BDI are identified: male sex, age＞60 years, gallbladder inflammation,history of previous percutaneous cholecystostomy, Mirizzi’s syndrome, and previous multiple episodes of cholangitis with a history of biliary stenting [16–19] . However, Gr?nroos et al. also reported that females are at higher risk of severe BDI [20] . In our audit, two female patients sustained BDI compared to one male patient.

Open conversion rates and intraoperative complications were higher in patients with difficult cholecystectomies, and this is not surprising. What is interesting though, is the observation that in the majority of instances (60%), the primary general surgeon decided for open conversion before seeking an on-table HPB consult.Due to the retrospective nature of the study, it is impossible to ascertain the exact sequence of timelines, and thus we remain cautious about concluding that primary general surgeons should await HPB surgeons to scrub up and decide if an open conversion is warranted. The World Society of Emergency Surgery has published a scoring system to predict open conversion based on intraoperative variables [21] . A general surgeon should be aware of these risk factors and resort to bail-out strategies or seek an HPB consult when encountered with difficult cholecystectomy. Our opinion is that the only pressing need for expeditious open conversion is torrential bleeding, and only two patients had torrential bleeding from arterial injury as an intraoperative event. Thus, it is possible that if an HPB surgeon was called prior to open conversion,HPB surgeon could have completed the minimal access cholecystectomy safely, considering bail-out strategies. However, this cannot be proven in this study. Our experience with subtotal cholecystectomy, cholecystectomy in patients with previous percutaneous cholecystostomy, and unit report on index admission cholecystectomy show low overall conversion rates [ 9 , 11 , 12 , 18 ]. As acute cholecystitis is a common indication for cholecystectomy, the use of the Tokyo Guidelines 2018 may be a useful risk stratification tool, where a general surgeon could have lower threshold for “call for help” in patients with grade II or higher cholecystitis [22] . Unfortunately, we cannot compare patients’ outcomes with proactive versus reactive referrals, as only one patient had a proactive referral. A randomized controlled trial would not be possible, as it would be unethical to withhold a consult when one is warranted.Thus, every piece of evidence regardless of its level of hierarchy should be reported.

Length of surgical training and apprenticeship is an essential factor that influences perioperative outcomes. Several studies examined the learning curve of surgeons in laparoscopic cholecystectomy, and it is estimated that 200 cases are required for a surgeon to be proficient in laparoscopic cholecystectomy [23] . Moore et al. demonstrated that the incidence of BDI during the 50th case is 0.17%, which is 10-fold lower than that of the first case [24] . Extrapolation of this evidence suggests that specialty training in HPB may improve surgical outcomes in laparoscopic cholecystectomy.This is further supported by Boddy et al., who reported that specializing in HPB and upper gastrointestinal (GI) surgery results in a lower incidence of BDI and intra-abdominal collection [25] . Hence,it is prudent for general surgeons to refer to HPB specialists to manage difficult cholecystectomy promptly. Seeking HPB consult in difficult cholecystectomy is keeping with good practice and advocacy for patient safety. Subtotal cholecystectomy is a common bailout strategy, and it is reported to have good clinical outcomes regardless of the fenestration or reconstituting techniques [26] . However, while we advocate good practice to seek early HPB consult in event of difficult cholecystectomy, we caution to conclude that laparoscopic cholecystectomy should be performed by HPB surgeons only. Studies have demonstrated the safety of laparoscopic cholecystectomy done by trainees compared to senior surgeons; for instance, Lavy et al. retrospectively reviewed patients who underwent elective laparoscopic cholecystectomy by a senior surgeon versus a resident: conversion rate, complication rate and mean length of stay were comparable, with longer operative time by resident compared to senior surgeon [27] . While Boddy et al. showed lower incidence of BDI when cholecystectomy was performed by upper GI surgeons compared to non-upper GI surgeons (0.1% vs.0.9%,P= 0.005), upper GI surgeons were more likely to use intraoperative cholangiogram (IOC) compared to non-upper GI surgeons (83.4% vs. 16.9%,P＜0.001) [25] . Because upper GI and HPB surgeons have more experience in laparoscopic cholecystectomy,which is a protective factor against BDI, fewer incidence of BDI may be confounded by use of IOC [28] . In view of the above reasons, we believe that HPB consult should be obtained early in the event of difficult cholecystectomy.

We believe that this study is important. Universally, it is advocated that in instances of intraoperative difficulty, “call for help”.However, there is severe paucity of reports on what happens after help arrives. It is not realistic to compare the handful of patients that require such intraoperative consult with vast majority of cholecystectomy patients where a surgeon is able to conclude surgery safely. Thus, an argument that our report excludes a comparator group does not reduce the impact of results.

There are a few limitations to this study. Firstly, this is a singleinstitution retrospective audit with a small number of patients, and thus results are not generalizable to the global surgical community.Each healthcare system needs to consider the resources and expertise available in treating patients with symptomatic gallbladder disorders. Secondly, outcomes of patients with reactive HPB consults may be intrinsically worse as patients are referred for difficult cholecystectomy or intraoperative complications. However, our results do suggest that HPB opinion could be sought prior to open conversion decision, at least in some patients. We also did not record the timing of referral to the HPB service, which has been shown to influence outcomes [2] . Thirdly, we did not report patients’ long-term outcomes with iatrogenic BDI as it would warrant a detailed medical record review and warrant an ethics clearance application with the local institutional review board under the new human biomedical research act. The new regulations mandate for investigators to obtain a “signed” consent copy from the patient or next-of-kin (if the patient has demised), and in our opinion,this is not warranted. Fourthly, we did not investigate the use of novel and emerging tools like indocyanine green dye fluorescent cholangiography in difficult cholecystectomies. A recent metaanalysis concluded that indocyanine green dye fluorescent cholangiography improves visualization of common hepatic duct [29] . It remains to be proven, if this indeed reduces BDI incidence.

In conclusion, there is paucity of data on what happens when an HPB surgeon is called for help. Calling for help in instances of BDI is obligatory; but in other instances it is a personal choice.Our study shows that calling for help prior to open conversion is lacking and this awareness should be raised in the general surgical community. On-table HPB consults remain an integral pillar of safe cholecystectomy culture. Whether surgical outcomes could be improved by early HPB surgeon engagement needs to be determined by larger multicenter investigations; further studies should also investigate if Tokyo Guidelines or difficult cholecystectomy scoring systems could be used to guide HPB specialist referrals.

Acknowledgments

None.

CRediT authorship contribution statement

Kai Siang Chan: Formal analysis, Resources, Software, Writing original draft. Elizabeth Hwang: Data curation, Formal analysis.Jee Keem Low: Data curation, Writing review & editing. Sameer P Junnarkar: Data curation, Writing review & editing. Cheong Wei Terence Huey: Data curation, Writing review & editing. Vishal G Shelat: Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing review & editing.

Funding

None.

Ethical approval

Not needed.

Competing interestNo benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.