A nomogram for predicting major gastrointestinal bleeding in patients treated with rivaroxaban

Rivaroxaban is a direct oral anticoagulant with the highest risk of anticoagulant-induced major gastrointestinal bleeding (MGIB). Currently, there is a lack of tools to identify patients at high risk of rivaroxaban-induced MGIB.

To establish a nomogram model to predict the risk of MGIB in patients receiving rivaroxaban.

Demographic information, comorbidities, concomitant medications, and laboratory test results were collected from 356 patients (178 diagnosed with MGIB) who were taking rivaroxaban between January 2013 and June 2021. Univariate and multivariate logistic regression analyses were used to identify the independent predictors of MGIB, and a nomogram was constructed based on these predictors. A receiver operating characteristic curve, Brier score, calibration plot, decision curve, and internal validation was used to evaluate the calibration, discrimination, and clinical usefulness of the nomogram.

Age, haemoglobin level, platelet count, creatinine level, prior peptic ulcer disease, prior bleeding, prior stroke, proton pump inhibitor use, and antiplatelet agent use were independent predictors of rivaroxaban-induced MGIB. These risk factors were used to establish the nomogram. The area under the curve of the nomogram was 0.833 (95%CI, 0.782-0.866), the Brier score was 0.171, the internal validation accuracy was 0.73, and the kappa value was 0.46.

The nomogram demonstrated good discrimination, calibration, and clinical applicability. Therefore, it could accurately predict the risk of MGIB in patients treated with rivaroxaban.