Dataset Information

Machine learning models compared to existing criteria for noninvasive prediction of endoscopic retrograde cholangiopancreatography-confirmed choledocholithiasis.

ABSTRACT:

Background and aims

Noninvasive predictors of choledocholithiasis have generally exhibited marginal performance characteristics. We aimed to identify noninvasive independent predictors of endoscopic retrograde cholangiopancreatography (ERCP)-confirmed choledocholithiasis and accordingly developed predictive machine learning models (MLMs).

Methods

Clinical data of consecutive patients undergoing first-ever ERCP for suspected choledocholithiasis from 2015-2019 were abstracted from a prospectively-maintained database. Multiple logistic regression was used to identify predictors of ERCP-confirmed choledocholithiasis. MLMs were then trained to predict ERCP-confirmed choledocholithiasis using pre-ERCP ultrasound (US) imaging only and separately using all available noninvasive imaging (US/CT/magnetic resonance cholangiopancreatography). The diagnostic performance of American Society for Gastrointestinal Endoscopy (ASGE) "high-likelihood" criteria was compared to MLMs.

Results

We identified 270 patients (mean age 46 years, 62.2% female, 73.7% Hispanic/Latino, 59% with noninvasive imaging positive for choledocholithiasis) with native papilla who underwent ERCP for suspected choledocholithiasis, of whom 230 (85.2%) were found to have ERCP-confirmed choledocholithiasis. Logistic regression identified choledocholithiasis on noninvasive imaging (odds ratio (OR) = 3.045, P = 0.004) and common bile duct (CBD) diameter on noninvasive imaging (OR=1.157, P = 0.011) as predictors of ERCP-confirmed choledocholithiasis. Among the various MLMs trained, the random forest-based MLM performed best; sensitivity was 61.4% and 77.3% and specificity was 100% and 75.0%, using US-only and using all available imaging, respectively. ASGE high-likelihood criteria demonstrated sensitivity of 90.9% and specificity of 25.0%; using cut-points achieving this specificity, MLMs achieved sensitivity up to 97.7%.

Conclusions

MLMs using age, sex, race, presence of diabetes, fever, body mass index (BMI), total bilirubin, maximum CBD diameter, and choledocholithiasis on pre-ERCP noninvasive imaging predict ERCP-confirmed choledocholithiasis with good sensitivity and specificity and outperform the ASGE criteria for patients with suspected choledocholithiasis.

SUBMITTER: Dalai C

PROVIDER: S-EPMC8855981 | biostudies-literature | 2021 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning models compared to existing criteria for noninvasive prediction of endoscopic retrograde cholangiopancreatography-confirmed choledocholithiasis.

Dalai Camellia C Azizian John J Trieu Harry H Rajan Anand A Chen Formosa F Dong Tien T Beaven Simon S Tabibian James H JH

Liver research (Beijing, China) 20211022 4

<h4>Background and aims</h4>Noninvasive predictors of choledocholithiasis have generally exhibited marginal performance characteristics. We aimed to identify noninvasive independent predictors of endoscopic retrograde cholangiopancreatography (ERCP)-confirmed choledocholithiasis and accordingly developed predictive machine learning models (MLMs).<h4>Methods</h4>Clinical data of consecutive patients undergoing first-ever ERCP for suspected choledocholithiasis from 2015-2019 were abstracted from a ...[more]

PMID: 35186364

Similar Datasets

Project description:Background & aimsCholecystectomy (CCY) after an episode of choledocholithiasis requiring endoscopic retrograde cholangiopancreatography (ERCP) with stone extraction reduces recurrent biliary events compared to expectant management. We studied practice patterns for performance of CCY after ERCP for choledocholithiasis using data from 3 large states and evaluated the effects of delaying CCY.MethodsWe conducted a retrospective cohort study using the ambulatory surgery, inpatient, and emergency department databases from the states of California (years 2009-2011), New York (2011-2013), and Florida (2012-2014). We collected data from 4516 patients hospitalized with choledocholithiasis who underwent ERCP. We compared outcomes of patients who underwent CCY at index admission (early CCY), elective CCY within 60 days of discharge (delayed CCY), or did not undergo CCY (no CCY), calculating rate of recurrent biliary events (defined as an emergency department visit or unplanned hospitalization due to symptomatic cholelithiasis, cholecystitis, choledocholithiasis, cholangitis, or biliary pancreatitis), mortality, and cost by CCY cohort. We also evaluated risk factors for not undergoing CCY. The primary outcome measure was the rate of recurrent biliary events in the 365 days after discharge from index admission.ResultsOf the patients who underwent ERCP for choledocholithiasis, 41.2% underwent early CCY, 10.9% underwent delayed CCY, and 48.0% underwent no CCY. Early CCY reduced relative risk of recurrent biliary events within 60 days by 92%, compared with delayed or no CCY (P < .001). After 60 days following discharge from index admission, patients with early CCY had an 87% lower risk of recurrent biliary events than patients with no CCY (P < .001) and patients with delayed CCY had an 88% lower risk of recurrent biliary events than patients with no CCY (P < .001). A strategy of delayed CCY performed on an outpatient basis was least costly. Performance of early CCY was inversely associated with low facility volume. Hispanic race, Asian race, Medicaid insurance, and no insurance associated inversely with performance of delayed CCY.ConclusionsIn a retrospective analysis of >4500 patients hospitalized with choledocholithiasis, we found that CCY was not performed after ERCP for almost half of the cases. Although early and delayed CCY equally reduce the risk of subsequent recurrent biliary events, patients are at 10-fold higher risk of recurrent biliary event while waiting for a delayed CCY compared with patients who underwent early CCY. Delayed CCY is a cost-effective strategy that must be balanced against the risk of loss to follow-up, particularly among patients who are ethnic minorities or have little or no health insurance.

Dataset Information

Machine learning models compared to existing criteria for noninvasive prediction of endoscopic retrograde cholangiopancreatography-confirmed choledocholithiasis.

Background and aims

Methods

Results

Conclusions

Publications

Machine learning models compared to existing criteria for noninvasive prediction of endoscopic retrograde cholangiopancreatography-confirmed choledocholithiasis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets