C-reactive protein to lymphocyte ratio combined with clinical features to construct a predictive model for upper gastrointestinal bleeding due to peptic ulcer.

Objective: This research aimed to determine the feasibility and accuracy of CLR and clinical features to formulate a prediction model for Peptic Ulcer (PU)-induced Upper Gastrointestinal Bleeding (UGIB).

Methods: The clinical data of 146 PU patients were prospectively collected, and patients were divided into the UGIB group (n = 48) and the non-UGIB group (n = 98). The factors affecting UGIB were analyzed using multifactorial logistic regression and collinearity analysis. The prediction model of UGIB was constructed, the predictive value of which was analyzed using the Receiver Operating Characteristic Curve (ROC) and Area Under the Curve (AUC), while the accuracy was analyzed using the calibration curve and Hosmer Lemeshow goodness-of-fit tests, and the application value was assessed using decision curve analysis (DCA).

Results: Statistical significance was observed between the two groups regarding HP infection, ulcer diameter, ulcer stage, use of nonsteroidal anti-inflammatory drugs, Neutrophil, LYM, NEUT/LYM Ratio (NLR), CRP, and CLR. HP infection, ulcer stage, use of NSAIDs, NLR, and CLR were independent risk factors for UGIB, and PCT was a non-independent risk factor. The AUC for this model was 0.921. The calibration curve of the model matched the actual curve. The model achieved a better fitting effect in predicting UGIB (χ = 8.5069, df = 8, p = 0.3856) and had a better clinical application value.

Conclusion: A predictive model for PU-induced UGIB, based on CLR and clinical features, can assist in developing clinical treatment plans to prevent UGIB.