Abstract
Background and Aims
Intraoperative cholangiography (IOC) is commonly performed during cholecystectomy for definitive testing of choledocholithiasis. The aim of the study was to evaluate the accuracy of IOC in hospitalized patients with suspected choledocholithiasis and outcomes of those receiving ERCP.
Methods
Consecutive hospitalized patients who received cholecystectomy with IOC at 2 centers were identified. Abnormal IOC (filling defect, meniscus sign, and/or absence of duodenal contrast) was diagnosed by the surgeon. Retained stones were assessed by definitive testing (MRCP, EUS, common bile duct exploration, and/or ERCP) performed after IOC.
Results
Of 847 patients (mean age, 44 ± 17 years; 624 women (74%)), 43 (5%) met American Society for Gastrointestinal Endoscopy high- and 665 (79%) intermediate-probability criteria for choledocholithiasis. Forty-two surgeons performed a median of 4 IOCs (range, 1-166); 254 patients (30%) had positive IOCs. During a follow-up of 21.1 ± 22.4 months, 257 patients (30%) received definitive testing, and 129 (15%) had retained stones. Sensitivity, specificity, positive predictive value, and negative predictive value of IOC for choledocholithiasis were 94.6% (95% confidence interval [CI], 89.1-97.8), 81.6% (95% CI, 78.6-84.4), 48.0% (95% CI, 44.1-52.0), and 98.8% (95% CI, 97.6-99.4), respectively. ERCP performed in 218 patients (26%) at a median of 1 day (interquartile range, 1-2) after IOC led to 18 adverse events (8%), including pancreatitis in 14 (6%).
Conclusions
Although IOC was highly sensitive for ruling out retained stones, specificity was modest, and more than half of the patients with abnormal IOCs would have received diagnostic ERCP without other testing. Given procedural risks, EUS or other less-invasive tests should be performed before ERCP in this population.
Intraoperative cholangiography (IOC) is commonly performed during cholecystectomy for the evaluation of occult or suspected choledocholithiasis.1 When an abnormality is detected on IOC suggestive of choledocholithiasis, ERCP is the primary therapy for clearance of the biliary tree because of the technical complexity associated with surgical common bile duct (CBD) exploration. Despite the high efficacy for complete extraction of bile duct stones, ERCP is associated with appreciable postprocedure adverse events, including pancreatitis in approximately 10% of patients.2 American Society for Gastrointestinal Endoscopy (ASGE) guidelines recommend IOC as a definitive testing option for evaluation of patients with intermediate probability for CBD stones based on biochemical testing and abdominal ultrasound (US).3 ERCP is subsequently recommended for patients with positive IOC found during cholecystectomy.
Previous studies evaluating the accuracy of IOC for assessing choledocholithiasis are inconsistent. Although a meta-analysis evaluating the accuracy of IOC demonstrated a high pooled sensitivity and specificity, similar to ERCP for evaluation of CBD stones,4 a number of studies showed modest specificity and positive predictive value (PPV), leading to the frequent performance of diagnostic ERCPs.5, 6, 7, 8, 9, 10 Large studies evaluating clinically relevant populations of hospitalized patients with suspected choledocholithiasis receiving cholecystectomy and IOC are lacking. Clarifying the accuracy of IOC among patients with suspected choledocholithiasis is important for optimal patient selection for ERCP to reduce postprocedural adverse events.
We hypothesized that the PPV of choledocholithiasis is suboptimal and that a substantial proportion of patients with a positive IOC do not have retained CBD stones at the time of ERCP. The aim of the study was to evaluate the diagnostic accuracy, sensitivity, specificity, PPV, and negative predictive value (NPV) of IOC for choledocholithiasis. We also examined the clinical outcomes of patients who underwent ERCP after IOC with or without retained choledocholithiasis.
Methods
Patient population
Institutional review board approval from the participating centers was obtained before initiating the study. Consecutive hospitalized patients aged >16 years who received cholecystectomy with IOC at Loma Linda University Health (January 2014 to June 2021) or Riverside University Health System (October 2016 to October 2021) were identified. Cholecystectomy was performed by the staff surgeon with or without a trainee from respective centers, and IOC was interpreted in real time by the operating surgeon. ERCP was performed by the same staff gastroenterologists with or without a trainee at both centers after reviewing available biochemical tests and IOC results per standard practice. EUS was performed by 7 of 10 endoscopists who performed ERCP and was available for part of the study period (January 2019 to October 2021) in one center (Riverside University Health System).
Subjects who met enrollment criteria were searched by using International Classification of Diseases, 9th and 10th revision codes (Supplementary Table 1, available online at www.igiejournal.org). Patients who received ERCP with sphincterotomy and/or placement of a biliary stent before cholecystectomy that may confound the accuracy of IOC were excluded. Furthermore, pregnant patients or those who received IOC for reasons other than suspected choledocholithiasis (ie, preoperative suspicion of pancreatobiliary neoplasm) were excluded.
Data
Demographic and clinical data including laboratory values on presentation and abdominal US findings (presence of cholelithiasis, choledocholithiasis, CBD diameter) were collected. Furthermore, cholecystectomy (laparoscopic or open approach), IOC (presence of filling defect, meniscus sign, or lack of contrast filling the duodenum), CBD exploration (presence of CBD stones and/or sludge visualized during extraction or choledochoscopy), ERCP (presence of CBD stones and/or sludge), MRCP (presence of CBD stones), and EUS findings (presence of CBD stones and/or sludge) were recorded. Finally, the clinical course during the index hospitalization including ERCP-associated adverse events (pancreatitis, bleeding, perforation) and length of stay as well as biliary events after the index hospitalization (hospitalization, diagnostic imaging, and ERCP) were carefully examined.
Study endpoints
The primary endpoint was the presence of choledocholithiasis on ≥1 of the definitive testing modalities (MRCP, EUS, ERCP, or CBD exploration) performed after IOC during the follow-up period. The primary analyses were accuracy, specificity, and PPV of an abnormal IOC for choledocholithiasis, given the main goal of reducing the number of diagnostic ERCPs. Secondary analyses were sensitivity, NPV, positive likelihood ratio, and negative likelihood ratio of IOC for choledocholithiasis. Additional analyses evaluating diagnostic properties of IOC for choledocholithiasis including bile duct sludge were performed. Finally, a subgroup analysis was performed for patients who met the clinical endpoint of receiving definitive testing or had greater than 6-month follow-up from IOC if definite testing was not performed to capture development of delayed symptoms after hospitalization.4
Definitions
High-, intermediate-, and low-probability biochemical and US criteria for choledocholithiasis on presentation were defined per the 2019 ASGE guidelines.3 A high probability of choledocholithiasis was defined by having choledocholithiasis on abdominal US or a bilirubin level >4 mg/dL plus a dilated CBD (≥6 mm) on abdominal US.11 Clinical cholangitis was excluded as high-probability criteria in this study given that biliary decompression is performed by ERCP rather than cholecystectomy as the initial treatment. Choledocholithiasis on abdominal US was defined by the presence of documented choledocholithiasis. An intermediate probability of CBD stones was defined by patients meeting at least 1 of the following criteria: abnormal liver biochemical tests, age >55 years, and dilated CBD on abdominal US.3 Abnormal liver test results were defined by levels that exceeded the hospital laboratory-defined upper limit of normal (bilirubin >1.2 mg/dL, aspartate aminotransferase >35 U/L, alanine aminotransferase >45 U/L, and/or alkaline phosphatase >110 U/L).
Choledocholithiasis was defined by the presence of a discrete stone documented on ≥1 definite tests (MRCP, EUS, ERCP, and/or CBD exploration). An abnormal IOC was defined by the presence of a filling defect, meniscus sign, and/or failure to drain contrast into the duodenum. CBD exploration was defined by the use of instruments (Fogarty catheter, basket, or choledochectomy) for a transcystic approach and performance of choledochotomy for a transcholedochal approach.12 Administration of intravenous glucagon or flushing of the CBD with saline solution alone was not considered CBD exploration.
A true-positive IOC was defined by a positive IOC and choledocholithiasis observed in at least 1 definitive testing. A false-positive IOC was defined by a positive IOC and no choledocholithiasis observed on any definitive testing or lack of biliary events during or after hospitalization. A true-negative IOC was defined as a negative IOC without development of biliary events during or after hospitalization. A false-negative IOC was defined as a negative IOC with choledocholithiasis observed on any definitive testing for development of biliary events during or after hospitalization. Severity of ERCP-associated adverse events including pancreatitis, bleeding, and perforation was categorized as mild, moderate, or severe.13
Statistics
Descriptive statistics are reported as mean ± standard deviation and median with range or interquartile range for continuous variables. Dichotomous variables are expressed as proportions with 95% confidence intervals (CIs). The Student t test was used to measure difference of continuous variables and the χ2 test for categorical variables. Univariate analysis was performed to identify predictors of the accuracy of IOC with a P < .10. Afterward, multivariate analysis was performed to identify independent factors associated with the accuracy of IOC. A 2-sided P < .05 was considered to be statistically significant. SAS software (SAS Institute Inc, Cary, NC, USA) was used for all analyses.
Results
Patient population
During the study period, 847 consecutive hospitalized patients at Loma Linda University Health (n = 419) or Riverside University Health System (n = 428) receiving cholecystectomy with IOC met the inclusion criteria. Mean patient age was 44 ± 17 years, and 624 (74%) were women (Table 1). On presentation, 283 patients (33%) had leukocytosis and 126 (15%) had gallstone pancreatitis. Based on presenting biochemical and US evaluations, 43 patients (5%), 665 patients (79%), and 139 patients (16%) met ASGE high-, intermediate-, and low-probability criteria for choledocholithiasis. Of the 43 patients who met high-probability criteria, 42 (98%) demonstrated decline or normalization of bilirubin and 18 (42%) received MRCP before cholecystectomy. Two patients (5%) who were found to have confirmed choledocholithiasis on MRCP proceeded with a surgical plan for cholecystectomy with CBD exploration. Of the 139 patients who met low-probability criteria, 45 (32%) had leukocytosis and 4 (3%) had pancreatitis.
Table 1.
Demographic data of 847 patients who received intraoperative cholangiography
| Characteristics | Value |
|---|---|
| Age, y | 44 ± 17 |
| Male ex | 223 (26) |
| Race/ethnicity | |
| Latino | 586 (69) |
| White | 162 (19) |
| African American | 56 (7) |
| Asian | 33 (4) |
| Other | 10 (1) |
| Presenting laboratory values | |
| White blood cell count >11.8 ×109/L | 283 (33) |
| Aspartate aminotransferase >35 U/L | 514 (61) |
| Alanine aminotransferase >45 U/L | 462 (55) |
| Bilirubin ≥4.0 g/dL | 61 (7) |
| Lipase >3 times upper limit of normal | 126 (15) |
| Abdominal US | |
| CBD ≥6 mm | 299 (35) |
| CBD stone | 14 (2) |
| Gallstones | 792 (94) |
| 2019 American Society for Gastrointestinal Endoscopy criteria∗ | |
| High | 43 (5) |
| Intermediate | 665 (79) |
| Low | 139 (16) |
Value are mean ± standard deviation or n (%).
CBD, Common bile duct.
Clinical cholangitis was not considered high probability for CBD stones.
Cholecystectomy was performed by 42 surgeons with a mean duration of 14.0 ± 12.3 years in practice who conducted a median of 4 IOCs (range, 1-166) during the study period. Laparoscopic surgery was performed in 826 patients (98%) and open surgery in 21 (2%) (Table 2).
Table 2.
Operative and postoperative clinical course
| Clinical course | Value |
|---|---|
| Cholecystectomy | |
| Laparoscopic | 826 (98) |
| Open | 21 (2) |
| Intraoperative findings | |
| Filling defect | 229 (27) |
| Meniscus sign | 5 (1) |
| Duodenal contrast | 116 (14) |
| Any abnormal findings | 254 (30) |
| Definitive testing after intraoperative cholangiography | 257 (30) |
| ERCP | 218 (26) |
| MRCP | 51 (6) |
| EUS | 23 (3) |
| Common bile duct exploration | 21 (2) |
Values are n (%).
During cholecystectomy, IOC demonstrated abnormal findings suspicious for choledocholithiasis in 254 patients (30%), including the presence of filling defects with or without a lack of duodenal contrast or meniscus sign in 229 (90%) and a lack of duodenal contrast with or without meniscus sign in 25 (10%). After IOC, 257 patients (30%) received additional definitive testing for choledocholithiasis including ERCP in 218 (26%), MRCP in 51 (6%), EUS in 23 (3%), and CBD exploration in 21 (2%) during a median follow-up of 14.5 months (range, 0-92.5) from IOC (Fig. 1).
Figure 1.
Flowchart of 847 patients receiving cholecystectomy with IOC. IOC, Intraoperative cholangiography.
Among 254 patients who had an abnormal IOC, 34 patients (13%) did not receive additional definitive testing given the clinical impression of passed stones based on resolution of biliary colic and improving liver tests, including 22 (9%) who had a greater than 6-month follow-up. Among 593 patients who had a normal IOC, 37 (6%) received additional definitive testing after developing abnormal liver tests or recurrent biliary pain, including 16 (3%) who required repeat hospitalization.
The prevalence of choledocholithiasis was 15%, which increased to 17% when sludge was also considered choledocholithiasis. The mean hospital stay was 4.7 ± 3.8 days during the index hospitalization including a mean hospital stay of 2.4 ± 3.0 days after cholecystectomy and IOC.
Diagnostic performance of IOC
Accuracy, sensitivity, specificity, PPV, and NPV of IOC findings suspicious for choledocholithiasis (presence of filling defect, meniscus sign, and/or failure to drain contrast into the duodenum) were 83.6% (95% CI, 80.9-86.0), 94.6% (95% CI, 89.1-97.8), 81.6% (95% CI, 78.6-84.4), 48.0% (95% CI, 44.1-52.0), and 98.8% (95% CI, 97.6-99.4), respectively (Table 3). Furthermore, accuracy, sensitivity, specificity, PPV, and NPV of IOC for choledocholithiasis including CBD sludge were 85.6% (95% CI, 83.1-87.9), 94.6% (95% CI, 89.6-97.6), 83.4% (95% CI, 80.7-86.4), 55.1% (95% CI, 50.8-59.3), and 98.7% (95% CI, 97.4-99.3), respectively. Finally, accuracy, sensitivity, specificity, PPV, and NPV of IOC demonstrating a filling defect alone as the only positive criteria for choledocholithiasis were 85.2% (95% CI, 82.7-87.6), 87.1% (95% CI, 80.4-92.2), 84.9% (95% CI, 82.0-87.4), 53.3% (95% CI, 48.6-57.9), and 97.1% (95% CI, 95.6%-98.1), respectively.
Table 3.
Diagnostic properties of intraoperative cholangiography for choledocholithiasis
| Choledocholithiasis alone (n =847) | 95% Confidence interval | Choledocholithiasis with or without sludge (n = 847) | 95% Confidence interval | |
|---|---|---|---|---|
| Prevalence, n (%) | 129 (15) | 147 (17) | ||
| Accuracy, % | 83.6 | 80.9-86.0 | 85.6 | 83.1-87.9 |
| Sensitivity, % | 94.6 | 89.1-97.8 | 94.6 | 89.6-97.6 |
| Specificity, % | 81.6 | 78.6-84.4 | 83.7 | 80.7-86.4 |
| Positive predictive value, % | 48.0 | 44.1-52.0 | 55.1 | 50.8-59.3 |
| Negative predictive value, % | 98.8 | 97.6-99.4 | 98.7 | 97.4-99.3 |
| Positive likelihood ratio | 5.1 | 4.4-6.0 | 5.8 | 4.9-6.9 |
| Negative likelihood ratio | .07 | .03-.14 | .06 | .03-.13 |
Of 847 patients, 624 (74%) received definitive testing for choledocholithiasis and/or had >6 months of follow-up from IOC. The accuracy, sensitivity, specificity, PPV, and NPV of IOC findings suspicious for choledocholithiasis were 79.8% (95% CI, 76.4-82.9), 94.6% (95% CI, 89.1-97.8), 76.0% (95% CI, 72.0-79.7), 50.6% (95% CI, 46.6-54.7), and 98.2% (95% CI, 96.3-99.1), respectively.
ERCP outcomes
After cholecystectomy, 218 patients (26%) received ERCP at a median of 1 day (interquartile range, 1-2), including 201 with abnormal and 17 with normal IOCs (Fig. 1). One patient (<1%) received a single-stage intraoperative ERCP for abnormal IOC findings. Of the 201 patients who received ERCP after abnormal IOC findings, diagnostic ERCP was performed in 80 patients (40%), which decreased to 64 (32%) when sludge was considered as choledocholithiasis. EUS was performed in 23 patients at a median 2 days (range, 1-141) after cholecystectomy, including 17 with abnormal and 6 with normal IOCs. Diagnostic ERCP was performed in 1 of 17 patients (6%) who received EUS and ERCP, whereas none of the 3 patients who received EUS without a confirmation of choledocholithiasis required an ERCP. Among 132 patients who had false-positive IOCs based on additional definitive testing, 80 (61%) received a diagnostic ERCP.
After ERCP, 18 patients (8%) had adverse events: pancreatitis in 14 (6%), perforation in 3 (1%), and bleeding in 1 (<1%) (Table 4). Of the 14 patients who developed pancreatitis, 1 was severe, 8 were moderate, and 5 were mild. Four patients who received diagnostic ERCP after false-positive IOCs had adverse events, including moderate pancreatitis in 3 and perforation in 1. One patient was found to have a Stapfer type II periampullary perforation that was recognized during ERCP and was managed by placement of a biliary stent and conservative therapy, requiring 9 days of hospitalization.
Table 4.
Adverse events after ERCP
| Patient no. | Intraoperative cholangiography results | ERCP findings | Adverse events | Severity |
|---|---|---|---|---|
| 1 | + | Stone | Pancreatitis | Mild |
| 2 | + | Stone | Pancreatitis | Mild |
| 3 | + | Stone | Pancreatitis | Mild |
| 4 | + | Stone | Pancreatitis | Mild |
| 5 | + | Stone | Pancreatitis | Mild |
| 6 | – | No stone | Pancreatitis | Moderate |
| 7∗ | + | No stone | Pancreatitis | Moderate |
| 8∗ | + | No stone | Pancreatitis | Moderate |
| 9 | – | Stone | Pancreatitis | Moderate |
| 10 | + | Stone | Pancreatitis | Moderate |
| 11 | + | Stone | Pancreatitis | Moderate |
| 12∗ | + | No stone | Pancreatitis | Moderate |
| 13 | + | Stone | Pancreatitis | Moderate |
| 14 | + | Stone | Pancreatitis | Severe |
| 15 | + | Stone | Perforation | Mild |
| 16∗ | + | No stone | Perforation | Moderate |
| 17 | + | Stone | Perforation | Moderate |
| 18 | + | Stone | Bleeding | Mild |
Patients who had ERCP for false-positive intraoperative cholangiography.
The mean hospital stay after surgery was increased by 1.9 days (95% CI, 1.4-2.3) in 218 patients receiving ERCP compared with 629 who did not receive ERCP. Specifically, the mean hospital stay was increased by 1.8 days (95% CI, 1.1-2.4) in 80 patients who had diagnostic ERCP for false-positive IOCs compared with 584 patients who had true-negative IOCs. Of the 218 patients who received ERCP during hospitalization, 37 (17%) required repeat ERCP at a median of 63 days (range, 1-292) for suspicion for incomplete extraction of stones in 28 (76%), bile leak in 4 (11%), management of ERCP-related perforation in 3 (8%), and failed cannulation in 2 (5%).
Predictors of accuracy of IOC
On univariate analysis, site of surgery (87% vs 80%, P = .02), CBD diameter ≥6 mm (81% vs 85%, P = .08), performance of preoperative definitive testing by MRCP and/or EUS (79% vs 85%, P = .03), and IOC accuracy rates of the surgeon stratified by quintile (71%-96%, P < .0001) were associated with accuracy of IOC for confirmed choledocholithiasis (Table 5). However, on multivariate analysis, only IOC accuracy rates by the surgeon comparing first (adjusted odds ratio, 10.96; 95% CI, 4.18-28.74), second (adjusted odds ratio, 3.81; 95% CI, 1.92-7.53), third (adjusted odds ratio, 1.86; 95% CI, 1.09-3.16), and fourth (adjusted odds ratio, 1.75; 95% CI, 1.03-2.97) with fifth quintiles were associated with accuracy of IOC for retained choledocholithiasis.
Table 5.
Predictors of accuracy of intraoperative cholangiography for choledocholithiasis
| Univariate |
Multivariate |
||||
|---|---|---|---|---|---|
| n/N | Percent | P value | Adjusted odds ratio (95% confidence interval) | P value | |
| Site of surgery and intraoperative cholangiography | |||||
| Loma Linda University Health System | 363/419 | 87 | .02 | 1.50 (.86-2.63) | .15 |
| Riverside University Health System | 345/428 | 81 | |||
| Age | |||||
| >55 y | 190/236 | 81 | .13 | ||
| <55 ye | 518/611 | 85 | |||
| Gender | |||||
| Male | 192/223 | 86 | .24 | ||
| Female | 516/624 | 83 | |||
| Bilirubin | |||||
| ≥4.0 g/dL | 50/61 | 82 | .72 | ||
| <4.0 g/dL | 658/786 | 84 | |||
| Common bile duct | |||||
| ≥6 mm | 241/299 | 81 | .08 | .75 (.51-1.11) | .15 |
| <6 mm | 467/548 | 85 | |||
| Pancreatitis | |||||
| Yes | 102/126 | 81 | .39 | ||
| No | 606/721 | 84 | |||
| Common bile duct stones | |||||
| Yes | 10/14 | 71 | .26 | ||
| No | 698/833 | 84 | |||
| Preoperative MRCP/EUS | |||||
| Yes | 172/218 | 79 | .03 | .75 (.50-1.15) | .19 |
| No | 536/629 | 85 | |||
| Years in surgical practice | |||||
| >10 y | 360/424 | 85 | .18 | ||
| 5-10 y | 153/179 | 85 | |||
| <5 y | 195/244 | 80 | |||
| Accuracy rate by surgeon | |||||
| 1st quintile | 128/134 | 96 | <.0001 | 10.96 (4.18-28.74) | <.0001 |
| 2nd quintile | 164/186 | 88 | 3.81 (1.92-7.53) | .0001 | |
| 3rd quintile | 176/210 | 84 | 1.86 (1.09-3.16) | .02 | |
| 4th quintile | 126/157 | 80 | 1.75 (1.03-2.97) | .04 | |
| 5th quintile | 114/160 | 71 | Reference | ||
Discussion
In this U.S. study of 847 hospitalized patients receiving cholecystectomy with IOC, 229 patients (27%) had abnormal IOC findings but only 129 (15%) had confirmed CBD stones on subsequent definitive testing. Specificity and PPV of IOC for choledocholithiasis were 81.6% (95% CI, 78.6-84.4) and 48.0% (95% CI, 44.1-52.0), respectively. After cholecystectomy, 218 patients received ERCP at a median of 1 day (interquartile range, 0-2) from IOC, and 18 (8%) developed adverse events, including pancreatitis in 14 (6%).
Gallstone disease is a common cause of hospitalization in the United States,14 and cholecystectomy is performed during the index hospitalization to reduce the morbidity of recurrent gallstone disease.3 Given occult choledocholithiasis in 10% to 15% of patients receiving cholecystectomy,15 some have advocated the use of routine IOC despite increased operative time and cost.16, 17, 18 Society of American Gastrointestinal and Endoscopic Surgeons guidelines provide conditional recommendations for liberal use of IOC during cholecystectomy with a low certainty of evidence.19 However, previous studies evaluating the accuracy of IOC for evaluation of choledocholithiasis are conflicting.5, 6, 7, 8, 9, 10 Despite controversy, IOC is frequently performed in practice. A survey study of 4100 U.S. general surgeons showed that 27% of respondents used routine IOC during cholecystectomy regardless of suspicion for choledocholithiasis.20 Clarifying the accuracy of IOC among patients with suspected choledocholithiasis is important for optimal patient selection for ERCP.
Consistent with our hypothesis, our study demonstrated a modest specificity of IOC of 81.6% (95% CI, 78.6-84.4) for choledocholithiasis that was further reduced to 76.0% (95% CI, 72.0-79.7) when only patients with a greater than 6-month follow-up were analyzed. More importantly, PPV was remarkably low at 48.0% (95% CI, 44.1-52.0), indicating that more than half of the patients with abnormal IOC results would receive a diagnostic ERCP in the absence of other definitive testing. A meta-analysis of 654 patients receiving IOC demonstrated a high pooled specificity (99%; 95% CI, 95-100) for choledocholithiasis.4 A prospective study of 1171 patients receiving routine IOC with cholecystectomy demonstrated a high PPV of 95%.18 However, suboptimal PPVs of IOC (25%-67%) for evaluation of choledocholithiasis have been reported among patients receiving postoperative ERCP.5, 6, 7, 8, 9, 10 For example, a study of 130 consecutive patients receiving ERCP at a mean of 7.0 ± 1.8 days for abnormal IOC findings demonstrated choledocholithiasis in 87 (67%).8 In a prospective study of 34 patients with abnormal IOCs who received postoperative EUS with or without ERCP by 2 independent endoscopists blinded to the results, 22 (65%) required stone extraction.9 However, EUS performed before ERCP demonstrated a higher PPV of 95% for choledocholithiasis. Finally, in a study of 5915 hospitalized patients receiving ERCP for abnormal IOC results during cholecystectomy identified from administrative data, only 1478 (25.0%) required stone extraction.10
Variable PPVs reported among studies may be related to differences in patient population and study design. Earlier studies demonstrating high PPVs were subject to bias because IOC was interpreted with the surgeon’s knowledge of the reference standard (ie, CBD exploration or intraoperative US) or with assistance of a radiologist.17,18,21,22 On the contrary, in studies evaluating the accuracy of abnormal IOCs in patients referred for EUS or ERCP, suboptimal PPVs were observed.5, 6, 7, 8, 9, 10 Potential discrepancies could also be explained by delays in endoscopic evaluation that may have allowed spontaneous passage of stones. However, in our study, with one of the lowest PPVs reported, ERCP was promptly performed at a median of 1 day after IOC. In clinical practice, the use of IOC is operator-dependent. In a meta-analysis of 8 randomized trials, technical success was achieved in 80% to 99% of IOCs performed at tertiary care centers. However, technical success was achieved in only 66% of the IOCs performed at a rural center.23
In addition, interpretation of IOC is not straightforward.24 Difficulties distinguishing choledocholithiasis from air bubbles with filling defects which are most specific for choledocholithiasis or impacted stones from papillary stenosis with failure to pass contrast into the duodenum during real-time cholangiography may have led to false-positive IOCs. In our study, more than a 10-fold difference in accuracy of IOC for choledocholithiasis was observed in the top compared with the bottom quintile stratified by surgeon accuracy rate. In a prospective study of 46 patients receiving IOC, serial cholangiograms demonstrated resolution of the filling defect observed at the index examination at 48 to 72 hours in 12 patients (26%), suggestive of a false-positive result.15 Finally, the definition of what constitutes a positive IOC has not been standardized, leading to subjectivity in the interpretation of results.25,26 With trends demonstrating a reduction in the numbers of IOCs and CBD explorations performed in the United States, goals to maintain high accuracy of IOC in clinical practice will be further challenged.27
In our study, IOC demonstrated high sensitivity (94.6%) and NPV (98.8%) for ruling out CBD stones. Our results are consistent with the studies demonstrating high sensitivity for ruling out choledocholithiasis in those who received selective (98%) or routine (97%) IOC.18,22
As expected, adverse events were frequently (8%) observed in patients who received ERCP. Although the incidence of post-ERCP pancreatitis in our study (6%) was within the range reported in literature, one study reported an increased risk of pancreatitis attributed to intraoperative injection of contrast during IOC.28 If an ERCP is performed, a single-stage intraoperative ERCP could be considered if logistically feasible. A meta-analysis of 8 randomized trials including 977 patients demonstrated that intraoperative ERCP is associated with reduced hospital stays and lower postoperative morbidity compared with postoperative ERCP.29 Specifically, ERCP performed using a laparoscopy-assisted rendezvous technique may reduce the risk of failed cannulation or procedure-related pancreatitis. Furthermore, 3 patients (2%) developed moderate pancreatitis and 1 developed perforation after receiving ERCP among those who had a false-positive IOC. In addition, patients who had a false-positive IOC required an increased hospital stay of nearly 2 days compared with those who had a true-negative IOC. Finally, a substantial portion of patients (17%) receiving ERCP after IOC required multiple procedures, mostly for management of biliary stents placed during the index procedure.
Our study has clinical implications. Given the high sensitivity and NPV, negative IOC findings can effectively rule out choledocholithiasis. However, given our results as well as others demonstrating modest specificity and marginal PPVs, selective rather than indiscriminate use of IOC during cholecystectomy may be warranted. If IOC demonstrates abnormalities, intraoperative US, EUS, or possibly MRCP should be performed to confirm the abnormal finding rather than proceeding directly to ERCP to avoid procedural risks (Fig. 2). At the national level, increased efforts to standardize the definition of abnormal IOC result and rigorous training to maintain a high level of competence for interpretation of IOCs are needed. Given the excellent sensitivity to rule out CBD stones coupled with data demonstrating superior cost-effectiveness of combined cholecystectomy and IOC as a single-stage intervention, IOC will continue to be an important diagnostic test for evaluation of choledocholithiasis.30
Figure 2.
Management algorithm for IOC performed for evaluation of choledocholithiasis. IOC, Intraoperative cholangiography; CBD, common bile duct.
The strengths of our study include the evaluation of accuracy of IOC performed by 42 surgeons with varying experience across 2 centers in a large number of patients with granular data, increasing the generalizability of our findings. Our study has limitations related to the retrospective study design. A complete follow-up of more than 6 months was not available in one-fourth of our patients, which may have affected the results. However, all patients were followed until discharge, and a subgroup analysis including those with sufficient follow-up demonstrated similar results.
In conclusion, IOC demonstrated high sensitivity for ruling out retained CBD stones. However, specificity was modest, and more than half of the patients with a positive IOC would have received diagnostic ERCP without other testing. Given the risk adverse events related to ERCP, EUS should be performed to confirm abnormal IOC findings before ERCP.
Disclosure
All authors disclosed no financial relationships.
Footnotes
How to cite iGIE articles: Johnson R, Webber C, Thompson TJ, et al. Article title. IGIE 2023;2:10-26.
Appendix
Supplementary Table 1.
International Classification of Diseases, 9th and 10th revision codes for cholecystectomy and intraoperative cholangiography
| Cholecystectomy | Intraoperative cholangiography | |
|---|---|---|
| International Classification of Diseases, 9th revision | 51.2, 51.21, 51.22, 51.23, and 51.24 | 87.53 |
| International Classification of Diseases, 10th revision | 0FT40ZZ, 0FT44ZZ, and 0FB48ZZ | BF030ZZ, BF031ZZ, BF03YZZ, BF0C0ZZ, BF0C1ZZ, BF0CYZZ, BF100ZZ, BF101ZZ, BF10YZZ, BF110ZZ, BF111ZZ, BF11YZZ, BF120ZZ, BF121ZZ, BF12YZZ, BF130ZZ, BF131ZZ, BF13YZZ, BF140ZZ, BF141ZZ, and BF14YZZ |
References
- 1.Hope W.W., Fanelli R., Walsh D.S., et al. Sages clinical spotlight review: intraoperative cholangiography. Surg Endosc. 2017;31:2007–2016. doi: 10.1007/s00464-016-5320-0. [DOI] [PubMed] [Google Scholar]
- 2.Kochar B., Akshintala V.S., Afghani E., et al. Incidence, severity, and mortality of post-ERCP pancreatitis: a systematic review by using randomized, controlled trials. Gastrointest Endosc. 2015;81:143–149. doi: 10.1016/j.gie.2014.06.045. [DOI] [PubMed] [Google Scholar]
- 3.ASGE Standards of Practice Committee. Buxbaum J.L., Abbas Fehmi S.M., Sultan S., et al. ASGE guideline on the role of endoscopy in the evaluation and management of choledocholithiasis. Gastrointest Endosc. 2019;89:1075–1105. doi: 10.1016/j.gie.2018.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gurusamy K.S., Giljaca V., Takwoingi Y., et al. Endoscopic retrograde cholangiopancreatography versus intraoperative cholangiography for diagnosis of common bile duct stones. Cochrane Database Syst Rev. 2015:CD010339. doi: 10.1002/14651858.CD010339.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Varadarajulu S., Eloubeidi M.A., Wilcox C.M., et al. Do all patients with abnormal intraoperative cholangiogram merit endoscopic retrograde cholangiopancreatography? Surg Endosc. 2006;20:801–805. doi: 10.1007/s00464-005-0479-9. [DOI] [PubMed] [Google Scholar]
- 6.Vadlamudi R., Conway J., Mishra G., et al. Identifying patients most likely to have a common bile duct stone after a positive intraoperative cholangiogram. Gastroenterol Hepatol. 2014;10:240–244. [PMC free article] [PubMed] [Google Scholar]
- 7.Spinn M.P., Wolf D.S., Verma D., et al. Prediction of which patients with an abnormal intraoperative cholangiogram will have a confirmed stone at ERCP. Dig Dis Sci. 2010;55:1479–1484. doi: 10.1007/s10620-009-0894-1. [DOI] [PubMed] [Google Scholar]
- 8.Bill J.G., Kushnir V.M., Mullady D.K., et al. Evaluation of patients with abnormalities on intraoperative cholangiogram: time to abandon endoscopic retrograde cholangiopancreatography as the initial follow-up study. Frontline Gastroenterol. 2016;7:105–109. doi: 10.1136/flgastro-2015-100597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Luthra A.K., Aggarwal V., Mishra G., et al. A prospective blinded study evaluating the role of endoscopic ultrasound before endoscopic retrograde cholangiopancreatography in the setting of “positive” intraoperative cholangiogram during cholecystectomy. Am Surg. 2016;4:343–347. [PubMed] [Google Scholar]
- 10.Gahagan J.V., Maximus S., Whealon M.D., et al. Analysis of endoscopic retrograde cholangiopancreatography after positive intraoperative cholangiogram: Is it necessary? Am Surg. 2016;82:985–988. [PubMed] [Google Scholar]
- 11.He H., Tan C., Wu J., et al. Accuracy of ASGE high-risk criteria in evaluation of patients with suspected common bile duct stones. Gastrointest Endosc. 2017;86:525–532. doi: 10.1016/j.gie.2017.01.039. [DOI] [PubMed] [Google Scholar]
- 12.Zerey M., Haggerty S., Richardson W., et al. Laparoscopic common bile duct exploration. Surg Endosc. 2018;32:2603–2612. doi: 10.1007/s00464-017-5991-1. [DOI] [PubMed] [Google Scholar]
- 13.ASGE Standards of Practice Committee. Chandrasekhara V., Khashab M.A., Muthusamy V.R., et al. Adverse events associated with ERCP. Gastrointest Endosc. 2017;85:32–47. doi: 10.1016/j.gie.2016.06.051. [DOI] [PubMed] [Google Scholar]
- 14.Peery A.F., Crockett S.D., Murphy C.C., et al. Burden and cost of gastrointestinal, liver, and pancreatic diseases in the United States: update 2018. Gastroenterology. 2019;156:254–272. doi: 10.1053/j.gastro.2018.08.063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Collins C., Maguire D., Ireland A., et al. A prospective study of common bile duct calculi in patients undergoing laparoscopic cholecystectomy: natural history of choledocholithiasis revisited. Ann Surg. 2004;239:28–33. doi: 10.1097/01.sla.0000103069.00170.9c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Photi E.S., El-Hadi A., Brown S., et al. The routine use of cholangiography for laparoscopic cholecystectomy in the modern era. JSLS. 2017;21 doi: 10.4293/JSLS.2017.00032. e2017.00032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Nickkholgh A., Soltaniyekta S., Kalbasi H. Routine versus selective intraoperative cholangiography during laparoscopic cholecystectomy: a survey of 2,130 patients undergoing laparoscopic cholecystectomy. Surg Endosc. 2006;20:868–874. doi: 10.1007/s00464-005-0425-x. [DOI] [PubMed] [Google Scholar]
- 18.Videhult P., Sandblom G., Rasmussen I.C. How reliable is intraoperative cholangiography as a method for detecting common bile duct stones? A prospective population-based study on 1171 patients. Surg Endosc. 2009;23:304–312. doi: 10.1007/s00464-008-9883-2. [DOI] [PubMed] [Google Scholar]
- 19.Brunt L.M., Deziel D.J., Telem D.A., et al. Safe cholecystectomy multi-society practice guideline and state of the art consensus conference on prevention of bile duct injury during cholecystectomy. Ann Surg. 2020;272:3–23. doi: 10.1097/SLA.0000000000003791. [DOI] [PubMed] [Google Scholar]
- 20.Massarweh N.N., Devlin A., Elrod J.A., et al. Surgeon knowledge, behavior, and opinions regarding intraoperative cholangiography. J Am Coll Surg. 2008;207:821–830. doi: 10.1016/j.jamcollsurg.2008.08.011. [DOI] [PubMed] [Google Scholar]
- 21.Wu S.C., Chen F.C., Lo C.J. Selective intraoperative cholangiography and single-stage management of common bile duct stone in laparoscopic cholecystectomy. World J Surg. 2005;29:1402–1408. doi: 10.1007/s00268-005-7694-3. [DOI] [PubMed] [Google Scholar]
- 22.Griniatsos J., Karvounis E., Isla A.M. Limitations of fluoroscopic intraoperative cholangiography in cases suggestive of choledocholithiasis. J Laparoendosc Adv Surg Tech A. 2005;15:312–317. doi: 10.1089/lap.2005.15.312. [DOI] [PubMed] [Google Scholar]
- 23.Ford J.A., Soop M., Du J., et al. Systematic review of intraoperative cholangiography in cholecystectomy. Br J Surg. 2012;99:160–167. doi: 10.1002/bjs.7809. [DOI] [PubMed] [Google Scholar]
- 24.Sanjay P., Tagolao S., Dirkzwager I., et al. A survey of the accuracy of interpretation of intraoperative cholangiograms. HPB. 2012;14:673–676. doi: 10.1111/j.1477-2574.2012.00501.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Gao H., Munasinghe C., Smith B., et al. What features on intraoperative cholangiogram predict endoscopic retrograde cholangiopancreatography outcome in patients post cholecystectomy? HPB. 2021;23:538–544. doi: 10.1016/j.hpb.2020.08.010. [DOI] [PubMed] [Google Scholar]
- 26.Ryberg A.A., Fitzgibbons R.J., Jr., Tseng A., et al. Abnormal cholangiograms during laparoscopic cholecystectomy. Is treatment always necessary? Surg Endosc. 1997;11:456–459. doi: 10.1007/s004649900389. [DOI] [PubMed] [Google Scholar]
- 27.Altieri M.S., Yang J., Obeid N., et al. Increasing bile duct injury and decreasing utilization of intraoperative cholangiogram and common bile duct exploration over 14 years: an analysis of outcomes in New York State. Surg Endosc. 2018;32:667–674. doi: 10.1007/s00464-017-5719-2. [DOI] [PubMed] [Google Scholar]
- 28.Sitaraman L.M., Knotts R.M., Kim J., et al. Increased risk of pancreatitis after endoscopic retrograde cholangiopancreatography following a positive intraoperative cholangiogram: a single-center experience. Clin Endosc. 2021;54:107–112. doi: 10.5946/ce.2020.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Liao Y., Cai Q., Zhang X., et al. Single-stage intraoperative ERCP combined with laparoscopic cholecystectomy versus preoperative ERCP Followed by laparoscopic cholecystectomy in the management of cholecystocholedocholithiasis: a meta-analysis of randomized trials. Medicine. 2022;101 doi: 10.1097/MD.0000000000029002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ali F.S., DaVee T., Bernstam E.V., et al. Cost-effectiveness analysis of optimal diagnostic strategy for patients with symptomatic cholelithiasis with intermediate probability for choledocholithiasis. Gastrointest Endosc. 2022;95:327–338. doi: 10.1016/j.gie.2021.08.024. [DOI] [PubMed] [Google Scholar]