As a high-frequency and highly sensitive examination method, IDUS has important clinical value in diagnosing the biliary tract and pancreatic duct stenosis. Some studies suggested that IDUS is more accurate than EUS, transpapillary biopsy, and cell brushing for detecting biliary tract malignant lesions (12, 13). IDUS is relatively sensitive and specific in differentiating cholangitis from early cholangiocarcinoma (14). Moreover, previous studies have suggested that IDUS increases the diagnostic sensitivity and the specificity for primary sclerosing cholangitis from 62.5 to 87.5% and 53.1 to 90.6%, respectively, and the diagnostic accuracy of malignant lesions from 55 to 90% (15). Because of the bile component in the bile duct as a medium, ultrasound has good conductivity. So far, several studies have reported on IDUS for the diagnosis of CBD stones (16, 17). For example, Kim et al. (18) found that IDUS is useful for detecting occult CBD stone on ERCP in icteric patients with highly suspected CBD stones. The results showed that stones or sludge were found in all cases, with an average diameter of 2.9 mm. Another study suggested that the sensitivity of ERCP for the diagnosis of stones was significantly affected by the size of stones; the diagnostic rate was decreased in patients with stone size below 8 mm. Endo et al. (19) suggested that if the diameter of the bile duct is larger than 12 mm, routine IDUS examination should be performed. However, there are no more reports about the reasonable intervention time of IDUS in stone extraction. On the one hand, the traditional ERCP can be smoothly achieved with the help of cholangiography, and IDUS is not required for the operation process. On the other hand, there is always a certain rate of residual stones in patients who undergo ERCP. In fact, early stone recurrence after ERCP is often associated with stone omission (8).

In this study, we applied IDUS before or after ERCP. IDUS performed before ERCP can provide additional information about the calculus. After lithotomy IDUS was applied as a double check method of complete clearance of stones. Our results showed that the fluoroscopy time of the IDUS-AL group was significantly lower compared to the IDUS-BL and IDUS-ALC groups due to the reduction of one post lithotomy fluoroscopy. However, the degree of complete stone clearance was not affected. IDUS before lithotomy can give a primary impression of the stones and is helpful for lithotomy. Yet, in this study, we found no significant difference in the accuracy rate for the estimation of stone size and number between the three groups. Therefore, whether IDUS should be applied as a routine examination before stone extraction needs to be further investigated. As a point of view, patients with high risk of stone residual should have IDUS confirmation.

The quality of cholangiography can be influenced by the concentration of contrast and the unintended air in CBD. The personal experience of the operator may also affect the quality of the image. The infusion of contrast into hilar is a challenging process, which can affect the detection of stones. Detection of stones behind the endoscopic shaft or at the end of CBD also depends on the operator’s rich experience. Theoretically, IDUS can show the stones behind the shaft and at the end of the CBD more clearly than radiography, but the parapapillary diverticulum can also present as occupying lesion image under IDUS. Although it is difficult to fill the hilar with contrast, the contrast defect does not hinder the judgment of calculus. Our results also showed that in RL cases, the location of the stones was in the middle or lower part of CBD.

Generally speaking, no residual stones were found during follow-up in the three groups, and no serious complications occurred. Compared with other researches, we achieved a lower complication rate, which can be partly attributed to the assistance of general anesthesia and proper indication for ERCP. The shorter average FT can be due to the personal experience of the operator. Moreover, RL during ERCP was not emphasized by other researchers. Our study showed that RL was related to the accuracy of stone prejudgment. Inaccurate prejudgment led to a higher RL rate. Also, both lithotripsy and bile duct diameter were relevant to RL. After lithotripsy, many stone fragments were produced, so it is difficult to judge whether all the stones were removed. Previous studies have suggested that biliary stenting after lithotripsy is safe, which is consistent with our experience (20). When the bile duct is too wide, it can lead to inadequate angiography, making the basket empty out and the balloon not fit the bile duct wall. The high sensitivity of IDUS makes it possible to detect stones 1–2 mm in size. We are also concerned about whether the stones removed in RL are larger than 3 mm, because stones larger than 3 mm might cannot be automatically discharged. The high sensitivity of IDUS may also lead to misjudgment. Thick bile, guidewire, blood clot, and bubbles need to be differentiated from stones (Figure 3).

A few studies suggested that IDUS can reduce the residual rate of stones after ERCP. The residual stones missed by balloon occlusion cholangiography can be washed out with saline (10, 21). This raises another question, i.e., after sphincterotomy, even if stones with a 2–3 mm diameter are missing, can these missing stones flow out with bile. However, another prospective study found that additional IDUS can significantly reduce the stone recurrence rate three years after sphincterotomy (22). The analysis of risk factors for stone recurrence showed that patients with non-calculus gallbladder had low stone recurrence rate due to the flushing effect of bile flow.

This study has a few limitations. (1) There was heterogeneity among total patients according to the exaction methods of stones. Differences in the application of basket and balloon might influence the operation time or FT. After RL, patients did not receive further confirmation of total clearance. (2) No control group (e.g., patients who did not receive IDUS) was selected. IDUS is a time-consuming procedure, although it affects the operation time, it does not rely on fluoroscopy, so it does not affect the fluoroscopy time. IDUS showed a high sensitivity for detecting calculus but was not very effective for quantitative analysis of the number of stones, which can be partially attributed to our score system. The grading system for stones was developed based on the difficulty of lithotomy and the length of operation time. Although it showed to be sufficient for the evaluation of stone clearance, it can cause misjudgment for IDUS.

It was found that no matter IDUS was performed before stone removal for pre-judgment, or IDUS evaluation after stone removal, complete stone removal was completed without complications. After stone removal, IDUS can alone be used to confirm whether the stone is completely removed, reducing the radiation time. In summary, although prejudgment can simplify the process of lithotomy, the time and way of using IDUS need to be considered. Our data suggested that IDUS examination around lithotomy is helpful for the complete and safe extraction of stones. IDUS can also reduce fluoroscopy time and the rate of RL. Since IDUS can detect minimal stone in CBD, it can be potentially used as a golden standard for a total clearance of stones.