Preoperative gastric retention in endoscopic retrograde cholangiopancreatography patients: Assessing risks and optimizing outcomes

Abstract

This article is a comment on the article by Jia et al, aiming at establishing a predictive model to predict the occurrence of preoperative gastric retention in endoscopic retrograde cholangiopancreatography preparation. We share our perspectives on this predictive model. First, further differentiation in predicting the severity of gastric retention could enhance clinical outcomes. Second, we ponder whether this predictive model can be generalized to predictions of gastric retention before various endoscopic procedures. Third, large datasets and prospective clinical validation are needed to improve the prediction model.

Key Words: Endoscopic retrograde cholangiopancreatography; Preoperative gastric retention; Prediction model; Adverse event; Operation safety; Comment

Core Tip: Jia et al conducted a retrospective analysis of patients undergoing endoscopic retrograde cholangiopancreatography preparation, identified factors influencing preoperative gastric retention, and developed a predictive model for its occurrence. We have three key points to share after reading this enlightening paper.

INTRODUCTION

Endoscopic retrograde cholangiopancreatography (ERCP) remains a pivotal diagnostic and therapeutic tool in the management of pancreatic and biliary diseases. ERCP-related adverse events primarily include post-ERCP pancreatitis, cholangitis, bleeding, and perforation[1]. Gastric retention can lead to restricted visibility, operational difficulties, and an increased risk of surgical complications. Preoperative gastric retention and its associated adverse events have not been adequately recognized during ERCP procedure.

COMMENT ON A RECENR PUBLISHED ARTICLE

We read with great interest an article titled “Establishment of predictive models and determinants of preoperative gastric retention in endoscopic retrograde cholangiopancreatography” by Jia et al[2]. They conducted a retrospective analysis on 190 patients who underwent ERCP preparation and analyzed factors influencing preoperative gastric retention in ERCP, revealing that factors such as gender, primary disease, jaundice, opioid use, and gastrointestinal obstruction contribute to the risk of gastric retention. Then they established a predictive model to predict the occurrence of preoperative gastric retention in ERCP. After reading this enlightening paper, we have three views to share.

First, further differentiation in predicting the severity of gastric retention could enhance clinical outcomes. In clinical practice, for patients with mild gastric retention, it is crucial to shorten the duration of the ERCP procedure, prevent intraoperative vomiting, and immediately halt the procedure in case of complications. For patients with severe gastric retention, the ERCP procedure should be terminated immediately. Fasting should be continued postoperatively, and a gastrointestinal decompression tube should be left in place with negative pressure applied to suction the gastric contents. ERCP should be performed again once no food residue can be suctioned out[3]. Due to the differing treatment measures, we believe that predicting various degrees of gastric retention would be more beneficial.

Second, we ponder whether this predictive model can be generalized to predictions of gastric retention before different endoscopic procedures. The generalization of a predictive model refers to its ability to perform well on new, unseen data that was not used during the training phase. In other words, a model with good generalization accurately predicts outcomes across different datasets, rather than just performing well on the data that it was trained on. Even though the dataset was based on ERCP patients, those predictive factors, which are gender, primary disease, jaundice, opioid use, and gastrointestinal obstruction, have minor connection with ERCP operation itself. The prediction model could be tested on different endoscopic procedures to test its generalization.

Third, large retrospective datasets and prospective clinical validation are needed to improve the prediction model. This single-center, retrospective dataset contains only 190 patients. The accessibility of retrospective data is uncontrollable, and the same indicator may not be judged consistently in retrospective data, limiting its clinical utility. Larger datasets from multiple medical centers may enhance the predictive ability by incorporating more patients and diverse data. Furthermore, prospective clinical trials can better validate its efficacy in real clinical scenarios, evaluating model ability by indicators judged consistently.

CONCLUSION

The work by Jia et al[2] provides an insightful method for predicting the occurrence of preoperative gastric retention in ERCP, helping us assess risks and optimize ERCP outcomes. The continuous improvement and validation of the predictive model in clinical practice are very important.