Our study demonstrates a deviation from the previous literature in that there appears to be an inferiority of MRCP in detecting choledocholithiasis in ASGE intermediate-risk patients in comparison with EUS and IOC. This is particularly important in that the ASGE intermediate-risk population is the group of patients that MRCP is used the most to help guide their management. In our study, MRCP was used ~ 40% of the time in the ASGE intermediate-risk group. This study suggests MRCP should be used less frequently due to inferior diagnostic accuracy when compared to IOC and EUS.

Of the 208 Pre-op MRCPs, 78 of were positive. Of these, 20/78 of the patient did not have subsequent identification of stones on ERCP or IOC. Likewise, there were many instances where pre-op MRCP was negative yet the patient had positive ERCP or IOC. 130/208 of the pre-op MRCP group had a negative test; however, 27/130 ultimately had positive stones on ERCP or IOC at a later date thus leading to the low sensitivity, specificity, positive, and negative predictive values reported in the results section.

There is conflicting data in the literature regarding the optimum management from a length-of-stay (LOS) perspective in the management of patients with suspected choledocholithiasis. Lin et al. demonstrated a reduced LOS with laparoscopic cholecystectomy and IOC of 3.9 days compared to MRCP (7.0 days) and pre-ERCP (6.9 days) [11]. Ward et al. demonstrated no difference in LOS [12].

There is also a cost burden to society to consider with each of the three tests and the subsequent management of patients. Unfortunately, there is also conflicting literature reports on the optimum management strategy from this perspective depending on what factors are included in the calculation as well as the costs associated within each healthcare system. The vast majority of patients undergoing MRCP ultimately receive an invasive test (either ERCP or lap cholecystectomy) subsequently in their admission [13] which raises the question on the cost-effectiveness of this approach. Many of the modalities recommended (EUS, IOC, MRCP) are based on availability and feasibility given individual hospital resources. Use of an available modality (e.g., MRCP) may be beneficial if the alternative leads to delays in transfer for a more advantageous modality for evaluation. One reason for the inferior performance of MRCP could be due to the challenge of detecting small calculi in the distal common bile duct using this modality. Others argue that MRCP may reduce the need for unnecessary ERCP [12] thus saving cost in terms of unnecessary procedures, anesthesia, and potential complications.

Routine IOC has been proposed as the most cost-effective approach [14, 15]. Performing an IOC, however, adds additional operating time and cost to laparoscopic cholecystectomy. There may be a lost “opportunity cost” to surgeons that is hard to characterize. In addition, patients may still require post-operative ERCP as many surgeons do not routinely perform common bile duct exploration. The discussion of IOC is also important in the context of the significant increase in robotic-assisted cholecystectomy (RAC). RAC is commonly associated with fluorescent indocyanine green (ICG) cholangiography. ICG cholangiograms, while very helpful in delineating biliary anatomy, do not provide adequate examination of the common bile duct for obstruction. Traditional fluoroscopic IOC can and should be safely performed when indicated during robotic-assisted cholecystectomy. While RAC is shown to have an overall higher rate of IOC with ICG compared to laparoscopic cholecystectomy (44.2% vs 1.4%), it is associated with lower rates of traditional fluoroscopic IOC (9.1% vs 4.8%) [16]. This may have implications particularly as it pertains to our education of surgical trainees and their experience with performing IOC in their future practice.

There are some important limitations to note with this study. First, while the data were obtained prospectively, it was observational in nature and non-randomized. Patients were treated by many different providers at 12 different centers each with different resources and algorithms for managing these patients. As expected, our study demonstrated a significant amount of heterogeneity in the management of these ASGE intermediate-risk patients. This likely introduces a significant amount of bias into our data. For example, many centers have varying levels of access to ERCP which would therefore change if a surgeon was more likely to obtain a pre-op ERCP versus proceeding to the OR for a cholecystectomy with IOC. While the data in this were obtained from 12 different centers, all of them were major academic centers. Therefore, there may also be biases in patient selection, management paradigms, and resources available that would render our data not applicable or reproducible in other hospital settings. Our study did not evaluate patient comorbidities, in hospital changes in patient status, or provider decision-making which would also influence management. Additionally, while the dataset does include CBD diameter from pre-operative imaging (US, CT), there were no data regarding the size of number of stones seen on pre-op, intra-op, or post-op imaging. There were also no data collected regarding the size of CBD, size of stones, or number of stones seen on IOC or ERCP. All of this is known to play a significant role in operative decision-making.

Another important point to note is that while EUS appeared to have the best diagnostic accuracy, these numbers were very limited and therefore there was a wider confidence interval. Furthermore, EUS is invasive and a highly operator-dependent technique that requires a provider with a very advanced skill set and thus may not be realistic in all care settings.

In summary, this study does not provide a clear answer to the question regarding optimal management of the intermediate-risk patients. As mentioned above, our dataset has significant limitations which in turn limits the conclusions one can draw from our research. It does, however, shed light on the continued importance of high quality and reliable intra-operative cholangiogram as patient of the management paradigm in these patients. Furthermore, we need to ensure our surgical trainees continue to have adequate training and experience performing IOC during minimally invasive cholecystectomy.