Intraoperative cholangiography 2020: Quo vadis? A systematic review of the liter

Konstntinos Georgiou , Griel Sndlom , Nihols Alexkis , Lrs Enohsson ,

a First Department of Propaedeutic Surgery, Hippokration General Hospital of Athens, Athens Medical School, National and Kapodistrian University of Athens, Athens 10679, Greece

b Department of Clinical Science and Education, Department of Surgery, Karolinska Institutet, Södersjukhuset, Stockholm 17177, SE, Sweden

c Department of Surgical and Perioperative Sciences, Surgery, Umeå University, Umeå 90187, SE, Sweden

Keywords: Cholangiography Cholecystectomy Bile duct stones Bile duct injury Cost

ABSTRACT Background: There are few randomized controlled trials with sufficient statistical power to assess the ef- fectiveness of intraoperative cholangiography (IOC) in the detection and treatment of common bile duct injury (BDI) or retained stones during cholecystectomy. The best evidence so far regarding IOC and re- duced morbidity related to BDI and retained common bile duct stones was derived from large population- based cohort studies. Population-based studies also have the advantage of reflecting the outcome of the procedure as it is practiced in the community at large. However, the outcomes of these population-based studies are conflicting. Data sources: A systematic literature search was conducted in 2020 to search for articles that contained the terms “bile duct injury”, “critical view of safety”, “bile duct imaging” or “retained stones” in combi- nation with IOC. All identified references were screened to select population-based studies and observa- tional studies from large centers where socioeconomic or geographical selections were assumed not to cause selection bias. Results: The search revealed 273 references. A total of 30 articles fulfilled the criteria for a large obser- vational study with minimal risk for selection bias. The majority suggested that IOC reduces morbidity associated with BDI and retained common bile duct stones. In the short term, IOC increases the cost of surgery. However, this is offset by reduced costs in the long run since BDI or retained stones detected during surgery are managed immediately. Conclusions: IOC reduces morbidity associated with BDI and retained common bile duct stones. The re- ports reviewed are derived from large, unselected populations, thereby providing a high external validity. However, more studies on routine and selective IOC with well-defined outcome measures and sufficient statistical power are needed.

Introduction

The value, cost effectiveness and diagnostic accuracy of intra- operative cholangiography (IOC) have been the subject of intense debate over the last two decades. Two large register studies in- cluding more than 50 0 0 0 patients indicate that IOC reduces the risk for bile duct injury (BDI) [ 1 , 2 ], a rare but devastating compli- cation of gallstone surgery. IOC provides a clear overview of the bile duct anatomy and the critical view of safety, i.e.,identification of the cystic duct before it is divided as well as it gives intraop- erative information regarding if there are bile duct stones present. However, others claim that careful dissection [ 3 , 4 ] or the use of near-infrared fluorescent cholangiography with indocyanine green IOC [ 5 , 6 ] is sufficient to adequately provide the critical view of safety and that IOC during routine laparoscopic cholecystectomy (LC) only prolongs procedure time but neither improves the rate of stone retainment nor reduces the rate of BDI.

One major argument for routine IOC is that common bile duct stones (CBDSs) may be detected leading to their intraoperative re- moval. Retained stones may cause pain attacks, cholangitis or bil- iary pancreatitis, so immediate clearance of bile duct stones should reduce morbidity [7] . On the other hand, as with all screening pro- cedures this may lead to overtreatment. Since the natural course of CBDSs leftinsituis not completely clear, it is impossible to analyze the value of their intraoperative extraction. It may be that asymp- tomatic CBDSs in patients undergoing cholecystectomy cause less harm than symptomatic ones, and that they would normally ei- ther pass spontaneously or remain asymptomatic in cases where they are not detected intraoperatively.

The lack of evidence and widely accepted guidelines regarding IOC have led to geographical differences in the routine practice of gallstone surgery. In some countries such as Sweden with a pop- ulation of 10.2 million citizens, routine IOC is considered an im- portant part of surgical safety [8] . In Greece, with a population of 10.7 million citizens, other aspects of the procedure are given higher priority. This is probably explained by the fact that young surgeons tend to follow the practices of their elder colleagues rather than seek evidence-based guidelines elsewhere. Conforming to local routines tends to strengthen divergences in the practice of cholecystectomy that is currently seen.

As yet, no published randomized controlled trial (RCT) has had sufficient statistical power to assess the effectiveness of IOC in pre- venting CBD injury or retained CBDSs, and the subsequent reduc- tion in related morbidities. The best evidence regarding the ad- vantages of IOC was derived from large population-based cohort studies. Though all studies based on patient registries and large cohorts are observational, they do have the advantage of providing sufficient power to analyze rare outcomes such as BDI and retained stones. Furthermore, they provide outcome data for healthcare sys- tems in different countries. However, it is impossible to perform comparisons between different countries as there are numerous factors apart from IOC that affect outcome.

The aim of the present review was to scrutinize the evidence regarding the impact of IOC on the rates of BDI and retained stones, and the cost effectiveness of IOC, from studies based on data from population-based registries and large cohort studies.

Methods

A systematic literature search was conducted in March 2020. The electronic database MEDLINE was searched according to the PRISMA protocol.

Medical Subject Heading (MeSH) terms and text words were used based on the following search strategy: (i) (bile duct injury) OR (critical view of safety) OR (bile duct imaging) OR (retained stones); (ii) (intraoperative AND cholangiography); (iii) #1 AND #2; (iv) only manuscripts with abstracts referring to humans, re- ported in English language and dated back to 10 years were in- cluded.

The publications that were found with this search strategy were screened to identify studies that fulfilled the criteria of being large unselected observational studies. The following criteria should thus be fulfilled: (i) single institution studies with at least 500 proce- dures, population-based cohort studies or register-based studies; (ii) no exclusion of patients based on aspects that could be related to the decision to perform IOC; (iii) data on IOC performance and data on BDI, management of bile duct stones or economy reported.

The search revealed 273 articles that could possibly meet the inclusion criteria. Two of the authors (Sandblom G and Enochsson L) independently screened the abstracts of the articles retrieved and judged their suitability. Any differences of opinion were re- solved by discussion. Sixty-three articles were found suitable for further examination ( Fig. 1 ). Thereafter, a full-text review of those articles was carried out.

Fig. 1. Identification of eligible studies from the literature databases.

Review and meta-analysis papers were excluded as well as RCTs, since they always have exclusion criteria. However, their ref- erence lists were used to retrieve any study with relevant data. Registry studies and large observational studies equivalent to regis- ter studies, were included in this systematic review. Finally, a total of 30 articles fulfilling the above criteria were included in the re- view.

Results

Study characteristics

Thirty full-text articles were identified, including 13 population- based studies (T able 1) [ 1, 2 , 9 –19] . Furthermore, since there are so few articles written on this subject, we also included single institution studies with more than 500 LC procedures included [20–33] (T able 2) as well as publications where the im- pact of IOC on LC outcome were given using data from local databases [34–36]. In one of these studies [35], the total number of procedures was not given, but since they had limited the study population to 332 patients undergoing LC for gallstone pancreatitis, the total database was assumed to include well over 500 patients.

Origin of the studies

The majority of population-based studies were from USA (7/13) [ 9 , 11–13 , 15 , 18 , 19 ], Sweden (5/13) [ 1 , 2 , 14 , 16 , 17 ] and one study from Switzerland [10] . Publications based on data from single institution databases or local databases were more geographically spread including South America [20] , North America [ 28 , 30 , 31 , 35 , 36 ], Asia [ 21 , 29 , 33 ], Europe [ 22–27 , 32 ] and Africa [34] .

IOC and prevention/detection of BDI

Table 1 Population-based studies on the impact of IOC on bile duct injury (I), bile duct stones (S) and total cost (C). Studies Type of study Country-setting No. of patients/procedures Time period Topic under evaluation Conclusion Remarks Törnqvist et al., 2012 [1] I Retrospective GallRiks Sweden 51 041 LC, 747 iatrogenic BDI identified May 1, 2005 to December 31, 2010 To examine whether BDI at LC affects survival and whether IOC prevents postoperative deaths Early detection of a BDI and the intention to use IOC are associated with improved survival after LC GallRiks is high—about 90% of cholecystectomies in Sweden are registered Törnqvist et al., 2015 [2] I Retrospective GallRiks Sweden 51 041 LC, 747 iatrogenic BDI identified. 9008 patients with acute cholecystitis May 1, 2005 to December 31, 2010 Patient- and procedure-related risk factors for BDI with a focus on the rate of intended IOC No preventive effect of IOC in uncomplicated gallstone disease. Operating in the presence or a history of acute cholecystitis and OC were significant risk factors for BDI. The intention to perform IOC was associated with a reduced risk of BDI in patients with concurrent or a history of acute cholecystitis Any proposed protective effect of IOC was restricted to patients with (or a history of) acute cholecystitis Altieri et al., 2018 [9] I Retrospective exam of a state-wide database USA 391 945 LCs 14 years Over the years, trend of IOC, BDI, complications and readmissions IOC rate significantly decreases from 12.4% to 10.4%, while BDI and complications increases Routine IOC at time of LC may help early identification and potentially decrease BDI rate Giger et al., 2011 [10] I Retrospective in a 114 institutions database Switzerland 31 838 patients 1995-2005 To identify BDI risk factors and assess the effect of IOC Male sex and prolonged LC are independent risk factors for BDI during LC. Frequent use of IOC does not reduce BDI Selective IOC, maybe worthwhile, but experience in its interpretation may be an increasing problem Halawani et al., 2016 [11] I, C Retrospective of ACS NSQIP USA 72 289 patients. Group 1: LC only (54 990 patients); Group 2: LC and IOC (17 299 patients) 2012 and 2013 Comparison of the 2 groups and data analysis of the following: 30-day mortality, readmission, return to OR and NSQIP collected morbidity IOC during LC is associated with a statistically significant decrease in re-admission rates, especially readmissions related to biliary complications 570 participating hospitals. The cost-benefit ratio of performing routine IOC vs. readmissions cost cannot be analyzed with the current data Lilley et al., 2017 [12] I Retrospective cohort study using nationwide inpatient Medicare claims data USA In 472 367 patients who had LC for gallstones, cholecystitis, cholangitis, or gallbladder obstruction, IOC was performed in 35% and BDI in 0.3% January 1, 2004 to December 31, 2011 Association between IOC and CBD injury during LC for non-neoplastic biliary disease and compare survival among those with or without CBD injury IOC during inpatient cholecystectomy is associated with increased incidence of CBD injury. CBD injury is associated with reduced survival after cholecystectomy Robust prospective studies are needed to determine which patients are most likely to benefit from IOC Mangieri et al., 2019 [13] I Retrospective of ACS NSQIP USA 217 776 patients. 166 102 (76%) LC were performed without IOC and the 51 674 (24%) cases used IOC. In 433 cases a BDI was identified with an incidence of 0.19% 2012-2016 To determine if the BDI rate with LC has now become equivalent or superior compared to the open era IOC does not protect from BDI, and cholecystitis continues to be a risk factor for BDI. When LC requires conversion to OC, BDI increases a hundred fold Of the 334 cases with BDI diagnosed during readmission, only 26% of the cases had an IOC performed during the index procedure Möller et al., 2014 [14] S Retrospective GallRiks Sweden 38 864 cholecystectomies of which 34 200 had IOC; 3828 patients with CBDS were analyzed May 1, 2005 to December 31, 2009 Strategies for handling CBDSs in terms of complication rates and/or incomplete clearance with need of intervention High rates of unfavorable outcomes are associated with no measures when CBDSs are found during LC For small stones, the choice of taking no measures is associated with a surprisingly high risk for unfavorable outcomes ( continued on next page )

Table 1 ( continued ) Studies Type of study Country-setting No. of patients/procedures Time period Topic under evaluation Conclusion Remarks Ragulin-Coyne et al., 2013 [15] I, C Retrospective in NIS data USA 111 815 non-weighted patients 2004-2009 Outcomes included BDI, complications, mortality, length of stay and cost Routine IOC does not decrease the rate of BDI but is associated with significant added cost. Surgeons’ routine use of IOC is correlated with increased rates of postsurgical procedures and increased complications NIS lacks certain clinical variables, including patient-level factors (e.g., assessment of disease severity, imaging, laboratory values), specific perioperative data and long-term follow-up /readmission information Rystedt and Montgomery, 2016 [16] I Retrospective GallRiks Sweden 55 134 LCs, 188 BDIs (0.3%) of which 64% returned SF-36. 2007-2011 Intraoperative detection of BDI on QoL QoL after BDI is comparable to uneventful LC if the injury is diagnosed intraoperatively. Immediate repair resulted in similar QoL as in the control group We suggest liberal use of IOC for early detection of BDI, and intraoperative repair whenever possible Rystedt et al., 2017 [17] I, C Retrospective GallRiks Sweden 55 134 LCs. EQ-5D was returned by 61% (312 of 512) of all participants; 64% of patients with a BDI (107 of 168), and 60% in the control group (205 of 347). 2007-2011 To examine if routine IOC during LC is a cost-effective approach for analyzing the cost of iatrogenic BDI Intraoperative detection with immediate intraoperative repair is the superior strategy with less than half the cost and superior functional patient outcomes than postoperative diagnosis and delayed repair. The cost per QALY is considered reasonable The calculated total cost for routine IOC hence is €386 and for on-demand IOC €607 Sheffield et al., 2012 [18] S, I, C Retrospective from Texas Hospital Inpatient Discharge Public Use Database USA 176 981 LCs from 212 hospitals 2001-2008 To evaluate the percentage of variance in the use of IOC that was attributable to patient, surgeon and hospital factors IOC use: from 2.4% to 98.4% of cases among surgeons and 3.7% to 94.8% of cases among hospitals. IOC use was associated with increased age, gallstone pancreatitis or CBD stones, Hispanic race, decreased illness severity, insurance and later year of LC Uncertainty regarding the benefit of IOC leads to wide variation in use across surgeons and hospitals Sheffield et al., 2013 [19] I Retrospective from Texas Hospital Inpatient Discharge Public Use Database USA From 92 932 LC patients, 37 533 (40.4%) had concurrent IOC and 280 (0.30%) had a BDI 2000-2009 When confounders were controlled, there was no statistically significant association between IOC and BDI. IOC is not effective as a preventive strategy against BDI during LC The cohort included Texas Medicare fee-for-service beneficiaries aged 66 years or older, while the majority of patients who undergo cholecystectomy were younger than 65 years IOC: intraoperative cholangiography; LC: laparoscopic cholecystectomy; BDI: bile duct injury; OR: operation room; CBD: common bile duct; OC: open cholecystectomy; CBDS: common bile duct stone; QoL: quality of life; QALY: quality-adjusted life year; ACS NSQIP: American College of Surgeons National Surgical Quality Improvement Program; NIS: Nationwide Inpatient Sample.

Table 2 Single institution or regional studies on the impact of IOC on bile duct injury (I), bile duct stones (S) and total cost (C). Studies Type of study Country-setting No. of patients/procedures Time period Topic under evaluation Conclusion Remarks Alvarez et al., 2014 [20] I Retrospective in a single institution database Argentina 11 423 routine IOC patients 1991-2012 BDI rate in IOC Routine IOC use is associated with low incidence of BDI and helps early detection It is almost impossible to perform an RCT with sufficient power to determine whether routine IOC can prevent BDI, and population-based studies remain the sole as important evidence to clarify this aspect Yousefpour Azary et al., 2011 [21] S Retrospective in a single institution database Iran 2060 patients 2006-2008 IOC accuracy during LC and its predictive power to detect bile duct abnormalities Routine use of IOC within LC seems reasonable and is recommended A normal IOC, almost always can prevent unnecessary postoperative ERCP Buddingh et al., 2011 [22] S, I Retrospective in a single institution database Netherlands Selective IOC in 421 patients (group A) versus 435 with routine IOC (group B) Group A: 2004-2006; Group B: 2007-2009 The impact of compulsory routine IOC in BDI The rate of major BDI is 1.9% in the selective IOC group compared to 0% in the routine IOC group Routine IOC policy has fewer major BDIs and better intraoperative stone management Giulea et al., 2016 [23] S Retrospective in a single institution database Romania 945 patients, 147 IOCs 2013-2014 (1 year) The value of IOC following pre and intraoperative criteria Criteria for selective IOC may significantly reduce the number of useless cholangiograms Criteria: history of jaundice, elevated ALP, GGT, AST, ALT and CBD size Hjartarson et al., 2016 [24] S Retrospective in a single institution database Iceland 920 MRCP patients. 7 had IOC 2008-2013 The MRCP ability to exclude choledocholitiasis in symptomatic patients MRCP can exclude choledocholitiasis and reduce the amount of unnecessary ERCP Use IOC as an alternative to ERCP Karvonen et al., 2011 [25] I Retrospective in a single institution database Finland 75 BDIs encountered in a total of 8349 (1616 OCs and 6733 LCs) procedures January 1997 to December 2007 BDIs sustained in LC vs. BDIs sustained in OC OC is associated with a high number of BDIs, if minor BDIs are included, which are missed at the time of surgery. More than 90% of Amsterdam types A, B, and C BDIs can be treated endoscopically Optimal BDI treatment requires a multidisciplinary team Nassar et al., 2016 [26] S Prospective UK 3635 patients 1992-2014 To evaluate the use and benefits of IOC during LC at a high-volume biliary surgery unit IOC can be safely and routinely performed in LC. It helps to identify CBD stones, delineate bile duct anatomy and facilitate single-stage management of CBD stones Routine IOC should be a main stay, in tandem with laparoscopic CBD exploration, in single-session management of CBD stones Photi et al., 2017 [27] S, I Retrospective in a single institution database UK 1005 LCs with routine IOC October 1, 2013 to September 30, 2015 CBD stone detection on IOC, CBD injury, complication rates, readmission rate, and mortality. Sensitivity, specificity, for detection of CBD stones on IOC from preoperative biochemistry and radiology CBD stone detection rate: 10.1%. Readmission rate: 0.03%. No reported CBD injuries and 1 mortality. Of the preoperative predictive factors investigated, the most specific for CBD stone detection: bilirubin at 89%. The most sensitive: preoperative MRCP at 77% Routine IOC is an effective method of detecting CBD stones and BDI with improved patient outcomes and health economics Riggle et al., 2015 [28] S Retrospective in a single institution database USA 668 patients. 38 with suspicion for retained stones had postoperative imaging and 22 (3.3%) had a retained CBD stones July 1, 2012 to December 31, 2013 To assess the need of image perioperatively to patients that are at high risk for retained CBD stones Patients with biochemical evidence of CBD stones who downtrend preoperatively can be safely managed by LC with omission of biliary tract imaging - ( continued on next page )

Table 2 ( continued ) Studies Type of study Country-setting No. of patients/procedures Time period Topic under evaluation Conclusion Remarks Shah and Shah, 2011 [29] S Retrospective in a single institution database Nepal A total of 68 (2.8%) IOCs were done in 2400 LC patients October 1, 2005 to September 31, 2009 To assess the need of better indicators for IOC in order to minimize unnecessary procedure and its morbidity Existing indications of IOC detects only few patients with CBD stones and has high false positive results. This increases bile duct exploration, cost and morbidity. There is need to redefine indications for IOC Existing protocol to select patients for IOC leads to far too many unnecessary IOCs, especially the “history” of pancreatitis Shawhan et al., 2015 [30] I Retrospective in a single institution database USA 96 IOC patients out of 1799 LCs September 2004 to September 2011 To determine the incidence of BL after selective IOC and identify risk factors for predicting BLs Neither patient factors nor postgraduate year level significantly increased the rate of leak, although surgery type (open and converted) was associated with significantly increased risk IOC is a safe procedure that can be performed by trainees at all levels. IOC is performed relatively infrequently as ERCP gains availability Sirinek et al., 2015 [31] S, I Retrospective in a single institution database USA 7427 LC patients January 1, 2004 to December 31, 2013 Rates of IOC, ERCP and MRCP use throughout the years A significant decrease in the use of IOC and preoperative ERCP in the last decade. However, 7.6% of patients still underwent an IOC in 2013. Use of IOC during LC is not obsolete and remains a valuable tool for the evaluation of bile duct anatomy, BDI or suspected choledocholitiasis Experience with IOC could be provided by instituting a policy of routine IOC in healthy patients undergoing elective LC for symptomatic cholelithiasis Thacoor et al., 2019 [32] S Retrospective in a single institution database UK 2215 LC patients. 113 had acute gallstone pancreatitis. Of these, 102 patients underwent LC and IOC October 1998 to December 2013 To assess if in patients with acute gallstone pancreatitis, IOC may reduce the need for preoperative MRCP Our results and long-term follow-up show that most patients presenting with acute gallstone pancreatitis can be safely and successfully managed with LC and IOC without requiring a preoperative MRCP As an MRCP before LC is rarely performed, we do not have large numbers to base our conclusions regarding MRCP. This is likely to bias our time-to-treatment measures Zang et al., 2016 [33] S Retrospective in a single institution database China 4240 LC patients. IOC group: 1972 LC patients and 213 had IOC. MRCP group: 2268 LC patients and 257 had MRCP January 2009 to December 2014 To evaluate if preoperative MRCP can safely replace IOC during elective LC in terms of retained CBD stones and BDI MRCP is an effective and safe strategy when conducting elective LC to treat gallstones. MRCP might replace IOC and allow LC to be performed safely LC with routine IOC is more cost-effective than universal MRCP and ERCP in excluding CBD stones Hamad et al., 2011 [34] S Retrospective in 4 institutions Egypt 2714 2004-2009 (6 years) Complications of LC without IOC Low complication rates when proper detection of patients with silent CBD stones and ERCP are available In included centers LCBDE is not a usual procedure for the management of CBD stones Johnson and Walsh, 2012 [35] S Retrospective in 11 institutions Canada 332 patients with pancreatitis January 1, 1997 to December 31, 2001 The impact of IOC on clinical outcomes after LC for gallstone pancreatitis IOC does not improve outcomes after LC for gallstone pancreatitis The reason why some patients had IOC and others did not is unknown Pham et al., 2016 [36] S Retrospective in 2 affiliated institution databases USA 520 patients. 246 at institution A (routine IOC) and 274 at institution B (restricted IOC) 2008 to 2015 Comparison of performing or not performing IOC Routine IOC is not necessary in the setting of mild gallstone pancreatitis with normalizing bilirubin Patients at institution B had a shorter duration of surgery, shorter LOS, and fewer postoperative ERCPs performed IOC: intraoperative cholangiography; BDI: bile duct injury; RCT: randomized controlled trial; LC: laparoscopic cholecystectomy; ERCP: endoscopic retrograde cholangiopancreatography; ALP: alkaline phosphatase; GGT: gamma- glutamyl transpeptidase; AST: aspartate aminotransferase; ALT: alanine aminotransferase; CBD: common bile duct; LCBDE: choledochoscopic common bile duct exploration; MRCP: magnetic resonance cholangiopancreatography; BL: bile leak.

The population-based publications focused on whether IOC is effective in early detection of BDI and reduction in morbidity aris- ing from BDI ( Table 1 ). Studies with this outcome produce some- what conflicting results. A study by Altieri et al. [9] , including nearly 400 000 LCs over a 14-year period, showed a relationship between a decrease in the utilization of IOC and an increase in BDI rate. In a retrospective study of 72 298 cholecystectomies from the American College of Surgeons National Surgical Quality Im- provement Program (ACS NSQIP) database, including 54 990 LCs without and 17 299 LCs with IOC, a statistically significant re- duction in readmission rates, especially for those related to bil- iary complications, was seen in the group where IOC had been performed [11] . Two studies on data from the Swedish chole- cystectomy quality registry (GallRiks) support the beneficial im- pact of IOC on the early detection and treatment of BDI leading to better functional outcome of patients as well as reduced to- tal cost compared to the situation where diagnosis and repair are delayed [ 16 , 17 ]. Another publication based on data from GallRiks also suggested that early detection of BDI and the intention to use IOC are associated with improved survival after LC [1] . However, the same group in a later study on 51 041 LCs from the Gall- Riks database, revised their initial conclusion applied to patients with concurrent or a history of acute cholecystitis only [2] , thus questioning the Swedish policy of routine IOC in every LC. Oth- ers have questioned the beneficial effects of IOC during LC. Lil- ley et al. [12] found that IOC during inpatient cholecystectomy was associated with increased incidence of CBD injury. However, as the authors speculate, this may be due to the selective use of IOC in patients at higher risk of CBD injury or as a diagnos- tic tool when BDI is suspected. Other authors have stated that IOCpersedoes not prevent or reduce BDI [ 10 , 13 , 19 ]. Those in favor of IOC often argue that even if IOC does not prevent BDI, the intraoperative detection of BDI is highly beneficial. This im- proves postoperative outcome and total cost, since early detec- tion enables immediate measures to limit the harm caused by the injury [ 1 , 16 , 17 ]. In a study from the Texas Hospital Inpatient Dis- charge Public database, including 176 981 LCs from 212 hospitals, Sheffield et al. found that uncertainty regarding the beneficial ef- fects of IOC led to a diverse practice of IOC among surgeons as well as between hospitals in the same county [18] . In Sweden, on the other hand, IOC is routine and considered crucial for safety. Studies based on the GallRiks database have pointed out the beneficial ef- fects of routine IOC in terms of early discovery of BDI and a reduc- tion in total cost [ 1 , 16 , 17 ]. Some, however, interpret the same data as providing support for a more selective approach [2] . There are also large studies based on single institution databases that have explored the impact of IOC on BDI [ 20 , 22 , 25 , 27 , 30 , 31 ]. However, since BDI occurs in only 0.3%-1.5% of cholecystectomies [37–40] , even randomized studies from single institution are generally too underpowered to draw any safe conclusion [5] .

Indocyanine green (ICG) dye fluorescent cholangiography is an alternative to IOC that may be as effective for visualizing the bile ducts and preventing BDI [41] . However, as ICG cholangiography is a relatively new technique, there are no reports that meet the inclusion criteria of the present review and present data on the safety of the technique.

IOC and detection of bile duct stones (BDSs)

Those in favor of IOC emphasize the advantage that BDSs are detected intraoperatively and extracted, if necessary or possible, at the same time. Others claim that a more expectant strategy is pre- ferrable since small stones often pass spontaneously [ 42 , 43 ]. How- ever, Möller et al. [14] in a study on 38 864 cholecystectomies reg- istered in GallRiks, where 34 200 (88%) had IOC of which 11.6% had BDSs, found less favorable outcomes to be associated with an expectant policy. The value of IOC in detecting stones is de- pendent, to some extent, on the treatment strategy of choice. In Sweden, for example, where routine IOC is practiced [8] , when a stone is detected intraoperatively, some form of active interven- tion to remove the stone is performed in 90% of cases [44] . Most studies on IOC and BDS detection, however, are from small single institution studies where it is difficult to draw any definite con- clusions. Some advocate routine IOC [ 21 , 22 , 26 , 31 ] whereas others prefer a more selective approach [ 23 , 29 ]. Those who advocate rou- tine IOC state that a non-negligible proportion of patients have un- detected stones, and that by performing routine IOC and immedi- ate stone clearance, postoperative complications such as pancre- atitis and cholangitis are avoided. On the other hand, those advo- cating a more selective policy claim that IOC exposes patients to unnecessary radiation, and that many IOCs can be avoided. There are also those who claim that a more thorough preoperative as- sessment of patients including magnetic resonance cholangiopan- creatography (MRCP) would remove the need for IOC [ 24 , 33 ]. Fi- nally, there are others who simply advocate that in patients with biochemical evidence of CBDS but who downtrend preoperatively, omission of IOC is the way to go [28] .

In summary, results of studies on IOC and BDS are conflicting. Many are small, underpowered, and retrospective, and it is diffi- cult to draw any safe conclusions. However, the largest population- based study published so far, including 38 864 patients regis- tered in the GallRiks database, suggests that conservative manage- ment of BDS increases postoperative complication rate [14] . Further prospective population-based randomized studies are warranted.

Economic aspects of IOC

Assessment of the economic impact of IOC is complicated by differences among the healthcare systems of countries and the routine practices of their hospitals. Short-term cost of materials used during IOC, prolonged operative time, salaries of operating staff must be included in the analysis, as well as the long-term expense of postoperative complications such as treatment of BDIs. In a study by Halawani et al.in 2016 [11] based on 72 289 patients undergoing LC with or without IOC, the use of IOC was associated with a significant decrease in readmission rates related to biliary complications. They found that IOC added around three minutes to operative time. They were unable to analyze the cost-benefits related to the additional cost of IOC versus the long-term cost of readmission due to biliary complications but suggested that the balance between short- and long-term costs should be the focus of future studies. On the other hand, Ragulin-Coyne et al. [15] , in a retrospective study on more than 110 0 0 0 patients, found that rou- tine IOC did not decrease the rate of BDI but was associated with a statistically significant increase ($930 per case) in cost. As the authors declared in this study, these data were from the Nation- wide Inpatient Sample (NIS) database that, in contrast to GallRiks, is an administrative database that lacks certain clinical data at the patient level such as image findings, laboratory values, operative time and long-term follow-up/readmission information.

In a Swedish study from 2017 on 168 BDI patients, the authors scrutinized the medical records as well as data from the GallRiks database and compared those with a control group without BDI. For long-term follow-up, an EQ-5D form was sent to all BDI and control patients. The authors showed that intraoperative detection and immediate repair resulted in a 50% reduction in total cost, irre- spective of the type of injury, compared to postoperative detection and repair [17] . They found that the cost per quality-adjusted life year (QALY) using routine IOC resulted in a reasonable incremental cost-effectiveness ratio (ICER), thus supporting routine IOC.

Discussion

The pros and cons of IOC remain the subject of debate. There are many factors included in the equation such as the bene- fit of intraoperative detection of unexpected BDI enabling im- mediate measures to reduce the potential harm caused by the injury [ 1 , 16 , 17 ]. This leads to significant reduction in cost due to fewer readmissions and costly interventions related to bile duct repair [ 11 , 17 ]. However, the opinion that IOC promotes early de- tection of BDI is questioned by others [ 10 , 12 , 13 , 19 ]. There are also conflicting opinions as to whether IOC should be routinely performed to detect and remove BDS intraoperatively. Some pro- mote an expectant policy regarding BDS, since they maintain that small stones are cleared spontaneously [ 42 , 43 ], whereas Möller et al. [14] in a population-based study found that leaving stones was associated with a less favorable outcome. The IOC debate should also consider the financial aspects, taking into account the struc- ture of the healthcare system of the country in question. In Swe- den, for example, IOC is routine during cholecystectomy. This en- ables intraoperative detection of BDS and extraction by intraopera- tive endoscopic retrograde cholangiopancreatography (ERCP) over a guidewire or intraoperative preparation prior to postoperative rendezvous ERCP [44] . The reason why intraoperative ERCP over a guidewire is commonly performed in Sweden is partly because it reduces the incidence of ERCP-related pancreatitis [45] but also be- cause intraoperative ERCP in Sweden is usually performed by the surgeon. In other countries such as Greece or USA, the majority of ERCPs are performed by a gastroenterologist who does not have the intraoperative support of a surgeon due to economic or profes- sional reasons. Furthermore, the financial situation of the health- care sector varies significantly among countries as well as regions therein. When resources are limited, the short-term cost of mate- rials used during IOC can have a much more important impact on the equation than long-term considerations such as postoperative complications and readmission rates.

Putting aside differences in resources, regional treatment tra- ditions and healthcare systems that may affect the surgeon’s at- titude towards IOC, conflict of opinions still occurs despite simi- lar healthcare background. In USA, for example, Ragulin-Coyne et al. [15] found that routine IOC did not decrease the rate of BDI but increased the cost, whereas Altieri et al. [9] came to the opposite conclusion.

Fortunately, BDI is a rare event, and with this in mind, fu- ture studies on the beneficial impact of routine or selective IOC should be designed with well-defined treatment failure parameters so that enough statistical power can be reached, and safe conclu- sions made.

In conclusion, the current status of the place of IOC in LC is an- alyzed regarding its role in preventing BDI and stone retention, and short- and long-term costs are discussed. Our conclusions from an- alyzing the to date published studies are that the majority of pub- lications indicate that IOC reduces morbidity associated with BDI and retained CBDSs. IOC in the short-term perspective increases the cost, while in the long-term perspective, it reduces costs since intraoperative BDIs are dealt with immediately.

Acknowledgments

We thank Professor Evangelos Georgiou, from Athens University, Greece, for invaluable advice and support.

CRediTauthorshipcontributionstatement

KonstantinosGeorgiou:Data curation, Writing – original draft.GabrielSandblom:Conceptualization, Formal analysis, Methodol- ogy, Validation, Writing – review & editing.NicholasAlexakis:Su- pervision, Validation.LarsEnochsson:Conceptualization, Formal analysis, Methodology, Project administration, Writing – original draft.

Funding

None.

Ethicalapproval

Not needed.

Competinginterest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the sub- ject of this article.