Development and validation of a mathematical model for predicting the development of gastro-oesophageal reflux disease based on oesophagogastroduodenoscopy

Abstract

The purpose of this study was to identify a set of prognostic factors for the progression of gastro-oesophageal reflux disease for use in the development of a mathematical model for predicting this disease based on the results of oesophagogastroduodenoscopy. The study identified a range of prognostic factors for gastro-oesophageal reflux disease and a statistical method was employed to determine the level of their correlation with the development of the disease. The study found a link between certain clinical indicators and the occurrence of gastro-oesophageal reflux disease, which led to the formation of a set of prognostic factors for the progression of gastro-oesophageal reflux disease, including heartburn, frequent belching, regurgitation, damage to the mucous membrane of the oesophagus, stomach, duodenum, the presence of chronic gastroduodenitis, gastrointestinal dysfunction, bile reflux. In creating the mathematical prediction model, the logistic regression method was used to identify the correlation between the patient’s clinical indicators and the occurrence of reflux disease and to determine the probability of its progression. To bring the clinical information in line with the statistical formula, it was assigned the values of independent variables, and the presence or absence of a particular indicator was coded using the binary number system. To test the developed model, recommendations were given to assess the statistical significance of the independent variables to determine its adequacy and to determine the predictive ability by testing on an independent sample of patients. The developed prognostic model is of great practical significance for patients, the healthcare industry, and the further development of the field, as it enables prompt detection of diseases and suitable prevention and treatment measures, increases the diagnostic potential of the industry, optimises the allocation of medical resources, and leverages machine learning and artificial intelligence capabilities based on the existing model