Development and Validation of a Modified, Simplified Prediction Model Based on the Kyoto Classification of Gastritis for Current Helicobacter Pylori Infection

Objective This study aims to validate the application value of the endoscopic Kyoto Classification of Gastritis for determining the current status of Helicobacter pylori (H. pylori) infection in a population, and to construct a modified simplified prediction model based on this scoring system. Methods Data were collected from 313 patients who underwent gastroscopy and a 13C- or 14C-breath test (UBT) at the Digestive Endoscopy Center of Chengdu Third People’s Hospital between June 2022 and July 2023, and from 175 patients at Qionglai Second People’s Hospital between April 2022 and July 2023. The dataset from our hospital was used as the development set to construct a simplified prediction model for the current H. pylori infection, while the external dataset was used for validation. The model was developed using binary logistic regression and clinical expertise, and ROC curve analysis and the DeLong test were employed to compare diagnostic performance. Results The Kyoto Classification showed area under the curve (AUC) values of 0.862 (95%, CI: 0.822–0.902) and 0.850 (95%, CI: 0.775–0.925) in the development and validation sets, respectively. The modified model incorporating the absence of a regular arrangement of collecting venules (RAC), mucosal swelling, and diffuse/spotty redness achieved higher AUC values: 0.922 (95% CI: 0.888–0.956) and 0.914 (95% CI: 0.864–0.964) (p < 0.05). Accuracy rates were 87.2% and 89.1% in the development and validation sets, respectively. Conclusion The modified, simplified prediction model demonstrated superior diagnostic performance to the Kyoto Classification. This makes it a practical tool for endoscopists to use when assessing current H. pylori infection.