Machine learning predictive models for acute pancreatitis: A systematic review

•What is already known on the topic•Machine learning is gradually being widely used in predicting acute pancreatitis.•No study has classified or summarised various prediction tasks for acute pancreatitis. The performance of models in different studies and the problems associated with model construction remain unclear.•What this study adds to our knowledge•Machine learning-based models have great predictive performance, and outperform conventional statistical models and clinical scores in some prediction tasks for acute pancreatitis.•The IJMEDI checklist is a new quality assessment tool, and scores can be attempted to be associated with it to evaluate the effects and reliability of machine learning-based models.•According to the IJMEDI checklist, there are some problems in the process of data collection, data preprocessing, and model validation in current studies on machine learning-based models for acute pancreatitis, which limit their widespread application in clinical practice. Acute pancreatitis (AP) is a common clinical pancreatic disease. Patients with different severity levels have different clinical outcomes. With the advantages of algorithms, machine learning (ML) has gradually emerged in the field of disease prediction, assisting doctors in decision-making. A systematic review was conducted using the PubMed, Web of Science, Scopus, and Embase databases, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Publication time was limited from inception to 29 May 2021. Studies that have used ML to establish predictive tools for AP were eligible for inclusion. Quality assessment of the included studies was conducted in accordance with the IJMEDI checklist. In this systematic review, 24 of 2,913 articles, with a total of 8,327 patients and 47 models, were included. The studies could be divided into five categories: 10 studies (42%) reported severity prediction; 10 studies (42%), complication prediction; 3 studies (13%), mortality prediction; 2 studies (8%), recurrence prediction; and 2 studies (8%), surgery timing prediction. ML showed great accuracy in several prediction tasks. However, most of the included studies were retrospective in nature, conducted at a single centre, based on database data, and lacked external validation. According to the IJMEDI checklist and our scoring criteria, two studies were considered to be of high quality. Most studies had an obvious bias in the quality of data preparation, validati