Elsevier

Gastrointestinal Endoscopy

Volume 86, Issue 3, September 2017, Pages 525-532
Gastrointestinal Endoscopy

Original article
Clinical endoscopy
Accuracy of ASGE high-risk criteria in evaluation of patients with suspected common bile duct stones

https://doi.org/10.1016/j.gie.2017.01.039Get rights and content

Background and Aims

ERCP is recommended for patients considered high risk for choledocholithiasis after biochemical testing and abdominal US. Our aim was to determine whether the American Society for Gastrointestinal Endoscopy (ASGE) guidelines accurately select patients for whom the risk of ERCP is justified.

Methods

Consecutive patients hospitalized with suspected choledocholithiasis at Sir Run Run Shaw Hospital who received biochemical testing, abdominal US, and definitive testing for choledocholithiasis (MRCP, EUS, ERCP, intraoperative cholangiogram, and/or common bile duct [CBD] exploration) were identified. Patients with choledocholithiasis on abdominal US, with bilirubin levels >4 mg/dL (normal values <1.2 mg/dL), bilirubin levels ≥1.8 mg/dL plus a dilated CBD and/or clinical cholangitis were considered high risk per ASGE guidelines.

Results

Of 2724 patients with suspected choledocholithiasis, 1171 (43%) met high-risk criteria. Definitive testing (MRCP in 2442 [90%], EUS in 67 [2%], ERCP in 659 [24%], intraoperative cholangiogram in 229 [8%], and CBD exploration in 447 [16%]) revealed choledocholithiasis in 1076 [40%] patients. The specificity of the ASGE high-risk criteria was 74% (95% confidence interval [CI], 72%-77%) and positive predictive value was 64% (95% CI, 61%-67%). Using a more restrictive criteria (choledocholithiasis on abdominal US, bilirubin >4 mg/dL plus dilated CBD) improved the specificity to 94% (95% CI, 93%-95%) and positive predictive value to 85% (95% CI, 82%-88%). Doubling or more of bilirubin to >4 mg/dL and ≥1.8 mg/dL at second testing had specificities of 98% (95% CI, 96%-99%) and 95% (95% CI, 93%-96%), with positive predictive values of 62% (95% CI, 48%-76%) and 54% (95% CI, 44%-65%), respectively.

Conclusions

Although ASGE high-risk criteria demonstrated >50% probability of the patient having choledocholithiasis, more than a third of the patients would receive diagnostic ERCPs. Criteria with choledocholithiasis on abdominal US and/or bilirubin levels >4 mg/dL plus a dilated CBD showed higher specificity and positive predictive value.

Section snippets

Study setting

This study was conducted at Sir Run Run Shaw Hospital, which is an urban tertiary-care medical center in east China, where approximately 1000 ERCPs and 3700 cholecystectomies are performed annually.

Study design

Before initiating this study, permission for research was obtained from the ethics committee at Zhejiang University, Sir Run Run Shaw Hospital. Consecutive patients hospitalized between January 2011 to December 2013 with suspected choledocholithiasis who received biochemical testing, abdominal US,

Patient characteristics

During the study period, 2724 patients with suspected choledocholithiasis met the study criteria. The mean (± standard deviation) age of the patients was 55 ± 15 years, 1179 (43%) were male, and 215 (8%) had prior cholecystectomy (Table 1). On presentation, 1878 (69%) patients had at least 1 abnormal liver test result, 1214 (45%) had abnormal bilirubin levels >1.2 mg/dL, 806 (30%) had levels ≥1.8 mg/dL, and 349 (13%) had levels >4 mg/dL. On abdominal US, 1419 (52%) patients had dilated

Discussion

Suspected choledocholithiasis is the most common indication for ERCP.10 Although highly effective for removal of choledocholithiasis, ERCP is a complex endoscopic procedure that often is associated with procedure-related adverse events and even death.10, 11 With the advent of other highly accurate diagnostic modalities to evaluate for choledocholithiasis such as MRCP and EUS, patient selection to minimize ERCP-associated adverse events has evolved to improve patient safety. Furthermore, adverse

Cited by (63)

  • Predictive models in EUS/ERCP

    2023, Best Practice and Research: Clinical Gastroenterology
View all citing articles on Scopus

DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.

See CME section; p. 559.

If you would like to chat with an author of this article, you may contact Dr Kim at [email protected].

View full text